Fate of SARS-CoV-2 coronavirus in wastewater treatment sludge during storage and thermophilic anaerobic digestion.

Since the COVID-19 outbreak has started in late 2019, SARS-CoV-2 has been widely detected in human stools and in urban wastewater. No infectious SARS-CoV-2 particles have been detected in raw wastewater until now, but it has been reported occasionally in human stools. This has raised questions on the fate of SARS-CoV-2 during wastewater treatment and notably in its end-product, wastewater treatment sludge, which is classically valorized by land spreading for agricultural amendment. In the present work, we focused on SARS-CoV-2 stability in wastewater treatment sludge, either during storage (4 °C, room temperature) or thermophilic anaerobic digestion (50 °C). Anaerobic digestion is one of the possible processes for sludge valorization. Experiments were conducted in laboratory pilots; SARS-CoV-2 detection was based on RT-quantitative PCR or RT-digital droplet PCR. In addition to SARS-CoV-2, Bovine Coronavirus (BCoV) particles were used as surrogate virus. The RNA from SARS-CoV-2 particles, inactivated or not, was close to the detection limit but stable in wastewater treatment sludge, over the whole duration of the assays at 4 °C (55 days) and at ambient temperature (∼20 °C, 25 days). By contrast, the RNA levels of BCoV and inactivated SARS-CoV-2 particles decreased rapidly during the thermophilic anaerobic digestion of wastewater treatment sludge lasting for 5 days, with final levels that were close to the detection limit. Although the particles' infectivity was not assessed, these results suggest that thermophilic anaerobic digestion is a suitable process for sludge sanitation, consistent with previous knowledge on other coronaviruses.

Introduction

SARS-CoV-2 RNA has been detected repeatedly by Reverse Transcription quantitative PCR (RT-qPCR) in human stools, irrespective of the clinical severity of the disease, and sometimes even after no more SARS-CoV-2 RNA was detected in the respiratory tract samples (reviewed in (Gupta et al., 2020; Trypsteen et al., 2020; Elsamadony et al., 2021)). In human stools, the detected concentrations are of the order of 103-107 SARS-CoV-2 genomes per mL (Foladori et al., 2020), and in a few cases, the presence of infectious SARS-CoV-2 particles has been reported (Wang et al., 2020; Xiao et al., 2020; Zhang et al., 2020).

Consistent with the presence of SARS-CoV-2 in human stools, SARS-CoV2-RNA has also been detected extensively in urban wastewater, in multiple countries worldwide (reviewed in (Mandal et al., 2020)), such as The Netherlands (Medema et al., 2020), Spain (Randazzo et al., 2020a, Randazzo et al., 2020b), Italy (La Rosa et al., 2020), the USA (Wu et al., Sherchan et al., 2020), Japan (Haramoto et al., 2020), Turkey (Kocamemi et al., 2020a, Kocamemi et al., 2020b), India (Kumar et al., 2021), Australia (Ahmed et al., 2020a, Ahmed et al., 2020b), France (Wurtzer et al., 2020) and the United Arab Emirates (Hasan et al., 2021). The detected concentrations are several orders of magnitude lower than in human stools, in the range of ∼1-3.103 SARS-CoV-2 genomes per mL (Foladori et al., 2020; Peccia et al., 2020a, Peccia et al., 2020b; Balboa et al., 2021; D'Aoust et al., 2021). The presence of SARS-CoV-2 RNA in raw wastewater gave rise to the development of wastewater-based epidemiology (WBE), for the early detection of COVID-19 transmission in communities. It also raised concerns about the fate of SARS-CoV-2 during wastewater treatment (WWT) and the possible associated risks (Amoah et al., 2020; Arslan et al., 2020; Foladori et al., 2020; Yang et al., 2020). Based on recent studies, the risks are actually probably very limited, since no infectious particles were detected in urban wastewater by Robinson et al. (2022). In Wurtzer et al. (2021), after the spiking of urban wastewater with high amounts (107-108 genomic units/L) of SARS-CoV-2 particles, the authors detected infectious SARS-CoV-2 particles, yet at levels which were at least 3 orders of magnitude lower than the total SARS-CoV-2 RNA levels, at moderate temperatures. The decay rate of SARS-CoV-2 in wastewater has been determined in several studies, with estimated T90 (time to achieve 90% (one log) reduction) values of 1.2 days at 24 °C and 5.5 days at 4 °C (de Oliveira et al., 2021) or 1.5 days at room temperature (Bivins et al., 2020), suggesting the rapid loss of infectivity of SARS-CoV-2 particles in such matrix.

The fate of SARS-CoV-2 during wastewater treatment has started to be addressed (Foladori et al., 2020; Brisolara et al., 2021; Hasan et al., 2021). Regarding WWT sludge, which is the focus of the present study, part of the sanitary concerns relates to the consequences of spreading on agricultural land, which is employed to valorize a significant proportion of WWT sludge. According to the situations, land spreading occurs after limited sludge treatment, such as solely storage, or after more advanced processing, such as thermal treatment or anaerobic digestion (AD), followed by composting. Moreover, according to the process line, different types of sludge can be mixed before treatment, typically primary sludge and activated sludge. AD is a classical process for WWT sludge treatment and valorization, notably thermophilic AD (usually 49–58 °C) as it is considered as an efficient process for sanitation purposes (Astals et al., 2012). Until now, only a handful studies have focused on SARS-CoV-2 in WWT sludge (reviewed in (Adelodun et al., 2022)). In particular, SARS-CoV-2 has been detected through its RNA in WWT primary sludge in Turkey (Kocamemi et al., 2020a, Kocamemi et al., 2020b), in the USA (Peccia et al., 2020a, Peccia et al., 2020b) and in Spain (Balboa et al., 2021), at levels rather higher than in the influent wastewater. Balboa et al. (2021) explained this latter observation by the high affinity of enveloped viruses for solids (Wellings et al., 1976; Ye et al., 2016). The reported concentrations vary from about 1 to 5.105 SARS-CoV-2 RNA copies per mL of sludge. Several authors proposed WWT sludge as a convenient epidemiological indicator, since in addition to the high SARS-CoV-2 RNA levels, WWT sludge may be less sensitive to rainfalls than influent wastewater (Peccia et al., 2020a, Peccia et al., 2020b; Balboa et al., 2021). In activated sludge, SARS-CoV-2 RNA was detected by Kocamemi et al. (2020) and in more than half of the samples by Serra-Compte et al. (2021); by contrast, it was barely detected by Balboa et al. (2021). In the same manner, no SARS-CoV-2 RNA was detected after thermal treatment and AD by Balboa et al.(2021). In Serra-Compte et al. (2021), SARS-CoV-2 RNA was detected after AD but not after AD combined to thermal hydrolysis. No information on the AD operating temperature is available in this latter article, but mesophilic conditions were presumably employed, since they are the most common. The effect of sludge AD on SARS-CoV-2 persistence was studied in laboratory reactors by Bardi and Oliaee (2021), using RT-qPCR for SARS-CoV-2 RNA detection. This latter work focused on the effect of the temperature and of the organic loading rate (OLR) on SARS-CoV-2 persistence. Both parameters negatively affected SARS-CoV-2 persistence and had moreover synergistic effects. The temperature is widely known to affect the stability of coronaviruses (Laude, 1981; Wang et al., 2005; Gundy et al., 2008) and in particular SARS-CoV-2 (Chin et al., 2020). SARS-CoV-2 particles could be relatively less sensitive to pH in a wide range of values (Chan et al., 2020; Chin et al., 2020).

In the present work, we studied the persistence of SARS-CoV-2 in urban WWT sludge at 4 °C and at room temperature (∼20 °C), to simulate storage at contrasting seasons, as well as during the thermophilic AD of WWT sludge (50 °C). We relied for this purpose on laboratory reactors and on SARS-CoV-2 RNA quantification by RT-qPCR or reverse transcription droplet digital polymerase chain reaction (RT-ddPCR). In addition to SARS-CoV-2, we used Bovine coronavirus (BCoV) as surrogate. At least one well-characterized surrogate for enveloped viruses is available, the bacteriophage Phi6 (family Cystoviridae) (Adcock et al., 2009; Turgeon et al., 2014; Aquino de Carvalho et al., 2017), whose interactions with sludge have recently been studied (Yang et al., 2022). However, numerous studies resorted to BCoV as a SARS-CoV-2 surrogate (Graham et al., 2021; LaTurner et al., 2021; Watanabe et al., 2022), due to their similarity and evolutionary proximity: both belong to the Betacoronavirus genus. It has thus been suggested that BCoV could be a good surrogate for SARS-CoV-2 (LaTurner et al., 2021). Before SARS-CoV-2 pandemics, BCoV had moreover already been used as a surrogate to evaluate the recovery yield of water concentration methods (Abd-Elmaksoud et al., 2014). To our knowledge, our study represents the first use of BCoV as a surrogate virus in sludge for persistence assays. It provides useful insights into SARS-CoV-2 genome stability in WWT sludge during storage and thermophilic anaerobic digestion, especially since studies on SARS-CoV-2 in sludge are presently in limited number.

Materials and methods

Strategy of sludge sampling in wastewater treatment plants

Two types of sludge samples were required for the study. Fresh urban sludge samples were used in all the assays. For the thermophilic anaerobic digestion (AD) assay, thermophilic digested sludge samples, serving as microbial inoculum, were added to fresh sludge. Each type of sludge was sampled in a distinct plant.

Overview of the two wastewater treatment plants

The sludge samples originate from two WWT plants (WWTPs) located in the north-west of Paris (France): the Seine Aval (SAV) plant, for fresh urban sludge samples, and Seine Grésillons (SEG) plant, for thermophilic digested sludge (Fig. 1 , Table 1 ). Both plants are supervised by the Parisian operator in charge of wastewater treatment, SIAAP (Syndicat Interdépartemental Pour l'Assainissement de l'Agglomération Parisienne). They are fed with wastewater from the same catchment: Paris city and downstream Paris conurbation. This catchment is characterized by a dense urbanization in an area of almost 730 km2 (118 municipalities including the city of Paris) and more than 6.1 million inhabitants, which induces a wastewater quality of mostly domestic type. SAV plant was selected because it is the biggest WWTP in the Parisian region, hence the most representative. In addition, it encompasses more various treatment processes compared to the other regional plants. SEG plant was selected as the only one to include a thermophilic AD process, among the plants operated by the SIAAP.

Fig. 1

Experimental design. SAV: Seine Aval plant – SEG: Seine Grésillons plant – SEC: Seine Centre plant – SEM: Seine Morée plant – MAV: Marne Aval plant – SAM: Seine Valenton plant.

Table 1

Characteristics of the two wastewater treatment plants selected for sludge sampling.

	SEG	SAV
Catchment area (km2)	225	730
Wastewater flow (m3/d)	300,000	1,500,000
Nominal Population Equivalent (PE)	1,000,000	6100 000
Treatment layout in nominal conditions	Pre-treatment - Physico-chemical lamellar settling - 3 stages biofiltration	Pre-treatment - Primary settling – 3 stages biofiltration - Membrane bioreactor (ultrafiltration)
Sludge anaerobic digestion process	Thermophilic (55°)	Mesophilic (37°)
Average hydraulic retention time (HRT)	15 days	20 days
Total solid (TS) produced by plant in 2020 (ton)	23,924	109,027
Sludge used in this study	Digested sludge	Fresh sludge [mixture of primary and biological sludge (30:70 to 40:60, v:v)]

Sampling of fresh and thermophilic digested WWT sludge

Sampling was performed at contrasting epidemiological situations described as “high” or “low” in the names given to the fresh sludge. The average number of new confirmed cases per day in France on the sampling dates is shown on the Supplementary Fig. S1 (data.gouv.fr).

Fresh sludge samples were collected two times from SAV plant, on the October 27, 2020 (noted as Sludge_high, Fig. 1, Table 2 ) and on the December 1, 2020 (Sludge_low, Fig. 1, Table 2). For the thermophilic AD assay, the thermophilic digested sludge was collected from SEG WWTP on the December 1, 2020 (Inoc, Fig. 1, Table 2), from the outlet of a thermophilic digester. The sludge usage is presented schematically in more details in the Supplementary Fig. S2.

Table 2

Physico-chemical characterization of the 3 sludge batches, ordered by sampling date.

Sample	Name	WWTP	TS (g/g, %)	Volatile solids (VS) (% TS)	pH	Used in	Spiking status
Fresh WWT sludgeOct. 27, 2020	Sludge_high	SAV	4.1	68.7	6.77	Persistence assay 2 (4 °C and RT, open and closed bottles)	No spiking
Fresh WWT sludgeDec 1, 2020	Sludge_low	SAV	3.3	78.7	6.7	Persistence assay 1 (4 °C and RT, closed bottles)Anaerobic digestion assay	Spiking
Thermophilic digested WWT sludge (inoculum)Dec 1, 2020	Inoc	SEG	2.8	60.1	–	Anaerobic digestion assay	–

Sampling was performed by manual collection in a 5 L clean polypropylene bottle and followed by immediate refrigeration. The samples were stored refrigerated until subsequent use, at most 30 h after the sampling.

Preparation of coronavirus particles and spiking of fresh WWT sludge samples

For all assays in this study (except those using Sludge_high in the 2nd persistence assay), part of the fresh sludge was spiked with particles of SARS-CoV-2, previously isolated from COVID-19 patients and/or of BCoV. The purpose of the spiking was to reach higher initial levels and thus to better analyze the decay. As it was not always possible to employ active SARS-CoV-2 particles for spiking during our study, due to security issues, we successfully used BCoV as a surrogate virus. The preparation of the viral pools used for spiking is described below.

SARS-CoV-2 positive pool preparation

About two dozen of human nasopharyngeal eluates of swab samples were selected, that had been previously tested positive for SARS-CoV-2 (RT-qPCR, CT values < 20). Those swab samples were pooled together, filtered on a 0.45 μm sterile PVDF membrane (Millipore) and stored at −20 °C until subsequent use.

After thawing and vortexing, the positive pool was divided into 2 aliquots. One of them was heated at 70 °C in a water bath during 15 min, to obtain an inactivated pool of SARS-CoV-2 positive eluate. The other one was used without any previous inactivation. SARS-CoV-2 pool was tested positive after inactivation, with CT values of 17 (N1) and 15 (N2), using 100 μL of the pool.

Bovine coronavirus (BCoV) pool preparation

BCoV was used as a surrogate virus. BCoV particles were harvested from a human ileocecal adenocarcinoma cell culture (HCT-8), resulting in a suspension with a titer of 105.1 DCP50/mL (dose that causes cytopathic effects in 50% of the cells). The suspension was aliquoted and stored at −80 °C until subsequent use. BCoV RNA concentration in the pool was of 6.93.108 genomic units (GU)/mL (see 2.3.2 section), corresponding to a CT of 23, using 100 μL of the pool.

Sludge spiking

For the first persistence assay, 6 batches of 100 g of fresh WWT sludge (Sludge_low) were spiked with either 100 μL of BCoV solution (2 batches) or 10 mL of SARS-CoV-2 aliquot (2 batches) or 10 mL of inactivated SARS-CoV-2 aliquot (2 batches). For the thermophilic AD assay, 4 batches of 30 mL fresh WWT sludge (Sludge_low) were each spiked with both 10 mL of inactivated SARS-CoV-2 aliquot and 100 μL of BCoV solution. BCoV solution and SARS-CoV-2 pools were vortexed before spiking. Sludge, once spiked, was vortexed as well for homogenization.

Moreover, 100 μL of the SARS-CoV-2 pool were stored at −20 °C, to test their positivity level (1st positive control). The remaining eluate from the pool of inactivated SARS-CoV-2 was stored at 4 °C in order to monitor its positivity during the first persistence assay (2nd positive control). For the BCoV pool, 100 μL of the suspension were stored at −20 °C, to test their positivity level (3rd positive control).

Laboratory sample processing and analysis

Except for the RT-ddPCR analysis, which was ensured by ID Solutions, molecular biology analyses were performed at LABOCEA laboratory. This latter is accredited by COFRAC for PCR analysis in animal health, and the analyses are performed under quality standards (ISO/IEC 17025: 2017); the molecular biology analyses follow the recommendation of the standard NF 47–600.

Sample processing and RNA extraction from sludge

Sludge processing before RNA extraction

One g of sludge was diluted in a 50 mL Falcon tube containing 10 mL of sterile PBS and 5 glass beads of 8 mm diameter. After vortexing during 1 min, 1 mL of the diluted sludge was collected in duplicates and centrifuged at 11,300 g for 5 min, in order to perform RNA extraction in duplicates.

Approximately 0.25 g of sterile zirconium powder was added on top of each pellet. LMAP™ lysis buffer (300 μL, Innovative diagnostic) was added before grinding 3 times during 1 min at 6.5 m s−1 with a FastPrep-24 homogenizer. The sample was then centrifuged during 2 min at 3000 g. The supernatant (∼100–150 μL) was transferred to a distinct sterile tube, for subsequent RNA extraction.

RNA extraction

An automated extraction on magnetic beads was performed. It was adapted from the ID Gene Mag Fast Extraction Kit (Innovative diagnostic) commercial protocol using a King Fisher Flex 96 (Thermofisher) equipment, with the following modification: 150 μL of LYS-FAST™ lysis buffer (Innovative diagnostic) were replaced by 300 μL of LMAP™ lysis buffer (Innovative diagnostic), more adapted to the sludge matrix.

For extraction, supernatant (100 μL), originating from a sample previously processed according to the procedure described in section 2.3.1.1, was distributed in a 96 deep-well plate. For each extraction series performed, 100 μL of PBS (the same used for fresh WWT sludge dilution) were also distributed in one well, as a negative control for RNA extraction. Moreover, 5 μL of HCT-8 cells were added to each well, to serve as extraction and inhibition controls.

Quantification of coronavirus RNA

RT-qPCR for the quantification of SARS-CoV-2 and BCoV RNA

For SARS-CoV-2, 5 μL of RNA eluate and 0.16 μL of bovine serum albumin were added to 8 μL of IDNCOV-2 master mix (Innovative diagnostic), relying on hydrolysis probes. The qPCR amplification was performed according to the manufacturer's instructions. Two specific regions of SARS-CoV-2 genome were targeted, N1 and N2, in the N gene, encoding the nucleocapsid protein. In addition, the human gapDH gene (encoding the glyceraldehyde 3-phosphate dehydrogenase in the HCT-8 cells) was targeted as a control for extraction and inhibition. Serial dilutions of synthetic nucleic acids were used as standards for absolute quantification via a standard curve (dilution factor of 10, from 1260 GU/μL to 0.126 GU/μL), from SARS-CoV-2 regions N1 and N2. Significant inhibition was considered to occur if a CT difference greater than 3 was observed between the sample and the negative control, for the human gapDH gene.

For BCoV, 5 μL of RNA eluate were added to 20 μL of VetMAX Ruminant Rotavirus & Coronavirus Kit (ThermoFisher Scientific) master mix, and the RT-qPCR was performed according to the manufacturer's instructions. The kit relies on hydrolysis probes. It also contains a passive internal reference, based on proprietary ROX dye. The N gene, which encodes the nucleocapsid protein, was targeted. Serial dilutions of synthetic nucleic acid from BCoV were used as a standard for absolute quantification via a standard curve (dilution factor of 5, from 20,000–32 GU/μL).

Either a Bio-rad Real time PCR CFX-96 apparatus with software Bio-Rad CFX Manager 3.1, or an Applied Biosystems 7500 apparatus with software 7500 Software v2.3 were used. Both apparatus were employed indifferently, as they yield similar results.

The RT-qPCR results were expressed in CT values in part of the present study. A CT value corresponds to the number of cycles required for the fluorescent signal to cross a defined threshold during the amplification. Therefore, higher CT values are associated to lower viral RNA levels, since a higher number of amplification cycles is required to reach the fluorescence threshold.

The limits of detection (LoD) by RT-qPCR were of 2.5 copies/μLand 7 copies/μL for SARS-CoV-2 and BCoV, respectively, according to the kit manufacturer. The limit of quantification (LoQ) was determined to be 4 copies/μL for SARS-CoV-2. Expressed in genomic units (GU)/g of total weight sludge, the LoD were equivalent to 2000 and 5600 for SARS-CoV-2 and BCoV, respectively, while the LoQ was equivalent to 4000 for SARS-CoV-2.

RT-ddPCR for the quantification of low levels of SARS-CoV-2 RNA

For the thermophilic AD assay, in order to improve the sensitivity of SARS-CoV-2 genome detection in samples potentially containing PCR inhibitors, we performed RT-ddPCR, using the IDNCOV-2d(s) kit (ID SOLUTIONS) on Naica Crystal digital PCR system (Naica Geode and Naica Prism; Stilla Technologies). It consists of an in vitro diagnostic test (Research Use Only) based on RT-digital PCR for the quantitative detection of SARS-CoV-2 viral RNA after RNA extraction. Primers and double-labeled probes (hydrolysis probe) were designed in three specific regions of the SARS-CoV-2 viral RNA: N1 and N2 from the N gene and IP2 from the RdRP gene (encoding the RNA-dependent RNA polymerase). The fluorescence signals were detected in the same FAM channel (green) to increase the sensitivity and specificity of the test: 3 droplets are theoretically emitted in the presence of one copy of SARS-CoV-2 genome. A human gene measured in the HEX channel (orange) was used as an internal control to verify the presence of the sample. The presence of nucleic acids was detected by an increase in fluorescence due to hydrolysis of the probes during amplification in each channel. PAC-NCOV-2 is the ready-to-use positive control included in the kit; it consists of a mixture of synthetic nucleic acids specific to SARS-CoV-2, and the human endogenous control. ARM-NCOV-2(s) is a ready-to-use reaction mixture containing Reverse Transcriptase, Taq polymerase, primers and hydrolysis probes. A positive control (PAC-NCOV2) and a negative amplification control (NAC) were systematically included in each series of analyses. Fifteen μL of ARM-NCOV-2 were added per well in addition to 10 μL of RNA extracted from each sample to be analyzed. The reaction conditions were in accordance with the manufacturer's recommendations.

According to the supplier, the Limit of Blank (LoB) values at 95% Confidence Interval of ARM-IDNCOV-2(s) and the Limit of Detection (LoD) are respectively of 2 and 3 droplets for the SARS-CoV-2 targets N1+N2+IP2.

Physico-chemical analysis

Total solids (TS), volatile solids (VS) and pH of the WWT sludge samples were determined. The analysis of TS was obtained after drying a sludge sample at 105 °C to eliminate the water content (NF EN 15934-A). The VS, usually used to evaluate the organic matter content, was determined by calcination at 550 °C (NF EN 15169). The pH was also monitored but only on fresh sludge (NF EN 15933), to design the experiments. All parameters were acquired by a certified laboratory of the SIAAP (COFRAC certification number: 1–1452, www.cofrac.fr).

Laboratory assays

Two persistence assays at low temperature (4 °C) and room temperature (20 °C), as well as one thermophilic AD assay (50 °C), were performed (Fig. 1). Due to the limited amount of material recovered from the swab samples, it was not possible to perform biological replicates in the present study. Technical duplicates were included.

Persistence assays

For all persistence assays, the monitored sludge batches were stored either in a cold room at 4 °C or at room temperature, at approximately 20 °C, in the laboratory. In the cold room, the temperature was monitored continuously with a temperature probe. The room temperature was monitored with a thermometer. Minimum and maximum temperatures registered by the thermometer were recorded daily during the experiment.

The first persistence assay (Supplementary Fig. S3) aimed at assessing SARS-CoV-2 persistence in fresh sludge at 4 °C or at room temperature. The fresh WWT sludge Sludge_low was employed (Table 2) in closed bottles. For each temperature condition (4 °C and room temperature), three aliquots were processed, of 100 g fresh WWT sludge each: they were spiked either with BCoV suspension (Sludge_low + BCoV), or with inactivated SARS-CoV-2 suspension (Sludge_low + SARS-CoV-2 inactivated), or with SARS-CoV-2 suspension (Sludge_low + SARS-CoV-2) (details on sludge spiking in section 2.2). At different time points for each temperature condition, 2 g of each spiked sludge were sampled. Among them, 1 g was stored at −20 °C for future assay, and 1 g was processed immediately (see section 2.3.1).

The second persistence assay (Supplementary Fig. S4) aimed at assessing the effect of closed versus open bottles on SARS-CoV-2 decay dynamics. The fresh WWT sludge Sludge_high was employed (Table 1). It had previously been tested positive for SARS-CoV-2 RNA (see result section 3.1), consistent with the epidemic context at the sampling date (Supplementary Fig. S1). The sludge batch was separated into 4 bottles, each containing 30 mL of sludge. Two bottles were stored at 4 °C and the other two bottles were kept at room temperature. For each temperature condition, one bottle was left open, and the other one was closed. At different time points for each temperature condition, 1 g of sludge was sampled and processed immediately (see section 2.3.1).

Thermophilic anaerobic digestion assay

This assay required to mix fresh sludge (Sludge_low), used as feeding, and digested sludge (Inoc), used as inoculum. An Automated Methane Potential Testing System (AMPTS II) (Supplementary Materials and Methods, Supplementary Fig. S5) was used to mimic the process of sludge AD. This batch AD system is usually employed to assess the methane potential of organic matters (André et al., 2016; Holliger et al., 2016). In the present study, methane production was monitored to control for the expected establishment and dynamics of the AD process, but will not be described.

Four 30 mL aliquots of fresh WWTP sludge (Sludge_low, Table 2) were spiked with 100 μL of BCoV suspension and 10 mL of inactivated SARS-CoV-2 suspension, and stored at 4 °C until the beginning of the AD assay, on the subsequent day. A schematic view of the AD assay in presented in the Supplementary Fig. S6.

Since it is was not possible to sample the liquid phase of reactors in the middle of the incubation period, due to the presence of tubing for gas collection (AMPTS system), selected reactors were sacrificed at desired time points in order to establish the viral decay dynamics. More precisely, monitoring of the AD kinetics was performed in 8 reactors, in thermophilic conditions (50 °C). Four of them contained 30 mL of spiked WWT sludge (S), 10 mL of water to rinse the bottle containing the sludge aliquot, and 70 mL of inoculum (Inoc, Table 2), sampled the day before, to provide a microbial consortium suitable for thermophilic digestion. The other remaining four reactors were similar except that the WWT sludge was not spiked (NS). The ratio employed between WWT sludge and inoculum had been previously identified as optimal to ensure a good AD process (André et al., 2016; Holliger et al., 2016). The viral decay dynamics was monitored by sacrificing pairs of reactors (S and NS) over a 5 days period, sufficient for most of the methane to be produced for the urban sludge (Yasui et al., 2008; Silvestre et al., 2011; Strömberg et al., 2014). The studied time points were day 0 (before the start of the digestion assay), and 1, 2 and 5.

After each sacrifice, 25 mL aliquots were prepared from the content of each reactor. The samples were stored at −80 °C and subsequently processed in less than 6 h (see section 2.3.1).

Results

SARS-CoV-2 levels in the studied sludge

Based on RT-ddPCR (Table 3 ), the thermophilic digested WWT sludge (Inoc) was negative for SARS-CoV-2. By contrast, the SARS-CoV-2 RNA level was of 1900 ± 620 GU/g in the sludge produced during a high epidemic level (Sludge_high, October 2020, 4 technical replicates), corresponding to 46,341 ± 15,122 GU/g TS. SARS-CoV-2 level was of 256 ± 0 GU/g in the sludge produced during a low epidemic level (Sludge_low, December 2020, 2 technical replicates), corresponding to 7758 ± 0 GU/g of TS. Finally, all sludge samples were tested negative for BCoV by RT-qPCR, before spiking.

Table 3

SARS-CoV2 and BCoV RNA levels in the WWT sludge samples.

Sample	Name	New confirmed cases per daya	SARS-CoV-2 levels			BCoV levels
			CT	GU/g	GU/g TS
Fresh WWT sludgeOct. 27, 2020	Sludge_high	44,744	35.2 (N1)34.6 (N2)	1900 ± 62	46,341 ± 15,122	<LoD
Fresh WWT sludgeDec 1, 2020	Sludge_low	10,045	37.6 (N1)36.7 (N2)	256 ± 0	7758 ± 0	<LoD
Thermophilic digested WWT sludgeDec 1, 2020	Inoc	10,045	<LoD			<LoD

Average number – CT: cycle threshold – GU/g: genomic units per gram of total weight – GU/g TS: genomic units per gram of total solids. LoD: Limit of Detection.

High persistence of SARS-CoV-2 in sludge at room temperature based on genome quantification

We led a persistence assay of SARS-CoV-2 at room temperature during 12 days and at 4 °C during 55 days, in laboratory conditions (Supplementary Fig. S3), using the fresh WWT sludge Sludge_low (Table 2, Table 3). The CT values obtained for the different samples ranging from 33.4 to 37.11, i.e. under the LoQ, it seemed more rigorous to express the results in CT (Fig. 2 ) than to convert them to GU/g of sludge. The CT values in the SARS-CoV-2 pool (Fig. 2, upper plot) were much lower than in the sludge (<20), confirming the high viral RNA concentration of the former. The pool was still positive after 157 days at 4 °C (Supplementary Fig. S7). Regarding the sludge spiked with BCoV or SARS-CoV-2 (Fig. 2, middle and bottom plots), no significant difference in RNA detection was observed during the 12 days of the room temperature assay. At the end of the 4 °C assay, at day 55, the RNA from both viruses was still detected.

Fig. 2

Persistence of BCoV and SARS-CoV-2 in spiked WWT sludge, at 4°C (55 days) and room temperature (12 days). Upper plot: CT values obtained for SARS-CoV-2 detection in the pool used for spiking (positive control). Middle plot: CT values obtained for SARS-CoV-2 detection in spiked sludge. Lower plot: CT values obtained for BCoV detection in spiked sludge. All CT values were obtained by RT-qPCR. The complete data corresponding to Fig. 2 are available in the Supplementary Table S1. In this assay, it was not possible to perform biological replicates due to the limited amount of SARS-CoV-2 pool available for spiking. Each curve therefore corresponds to the kinetics for a single pool or sludge batch.

In this first persistence assay, the laboratory reactors were closed, promoting anaerobic conditions. In real sludge storage conditions, there is usually no strict anaerobiosis. We therefore led a second persistence assay using Sludge_high without spiking (since it was naturally positive for SARS-CoV-2), to evaluate the effect of closing the reactors or letting them open. No major effect was observed (Fig. 3 ).

Fig. 3

Persistence of SARS-CoV-2 at 4°C and room temperature, in open (17 days) or closed (25 days) bottles, in unspiked fresh WWT sludge (Sludge_high). SARS-CoV-2 RNA levels were measured by RT-qPCR. Each curve corresponds to the kinetics of a single reactor.

The two distinct persistence assays show qualitatively similar trends, for both monitored viruses (SARS-CoV-2 and BCoV as a surrogate). The results overall support a rather high stability of SARS-CoV-2 and BCoV RNA in WWT sludge at 4 °C and ambient temperature. It does not mean that infectious viral particles are still present. It has indeed been shown for a different RNA virus, Poliovirus 1 (Picornaviridae family), which is non-enveloped, that the viral RNA can persist for much longer time periods than the infectious particles and that its decay pattern is different from the exponential decay of infectious particles (Gassilloud et al., 2003). The longer persistence of viral RNA compared to infectious particles has also been reported for SARS-CoV-1 by (Wang et al., 2005) and for SARS-CoV-2 in various aqueous matrices (Bivins et al., 2020; Sala-Comorera et al., 2021), including wastewater (Bivins et al., 2020).

No major difference in persistence was observed between SARS-CoV-2 and inactivated SARS-CoV-2. This result was unexpected since SARS-CoV-2 particles may have been partially altered by the inactivating thermic treatment, and could presumably present a lower persistence. We can hypothesize that the SARS-CoV-2 pool (not inactivated) was altered at a similar level due to the freeze-thaw step.

Fast decay of SARS-CoV-2 during thermophilic anaerobic digestion of sludge

To finely assess the decay dynamics of SARS-CoV-2 during AD, laboratory batch reactors were employed to mimic the digestion process in thermophilic conditions (50 °C) (Supplementary Fig. S6). We used a same batch of fresh sludge for the whole assay, Sludge_low (Table 2, Table 3). In addition to monitoring the naturally occurring levels of SARS-CoV-2 RNA in raw sludge, we also used sludge spiked with inactivated SARS-CoV-2 particles, in order to reach higher initial levels and thus to better analyze the decay. Finally, since spiking with active SARS-CoV-2 was not performed for security reasons, we also spiked the sludge with active BCoV particles, employed as a surrogate coronavirus, in addition to inactivated SARS-CoV-2. The viral RNA was quantified over time, by sacrificing a pair of reactors at each selected time point, one with unspiked sludge and one with spiked sludge.

During this thermophilic AD assays, it appeared that a high sensitivity was required to quantify SARS-CoV-2 RNA in the sludge matrix. Indeed, when using RT-qPCR, the signal was systematically below the quantification threshold, even at day 0 and after spiking (data not shown). We therefore turned to RT-ddPCR to reach a higher sensitivity (Fig. 4 A). By contrast, the RT-qPCR method was sufficiently sensitive for BCoV monitoring. The effect of the spiking was visible despite low levels, since 1620 GU/g of total weight were detected in average for SARS-CoV-2 at day 0, in case of spiking, compared to 532 GU/g of total weight in the absence of spiking. In both cases, the levels were the highest at day 0. In the subsequent days, they dropped to levels that were close to the detection limit, resulting either in the lack of detection, or to very low detected levels. At the end of the assay, at day 5, positive detection still occurred in one of the technical replicates in the raw sludge (without spiking), as well as in one of the replicates in the case of spiking (288 and 152 GU/g of total weight, respectively).

Fig. 4

Dynamics of SARS-CoV-2 and BCoV RNA levels during thermophilic anaerobic digestion of spiked or unspiked fresh WWT sludge. GU/g of total weight: genomic units per gram of total weight of samples from the anaerobic digestion batch reactors. The same batch of fresh WWT sludge was used for all assays, Sludge_low (Table 1, Table 2). Due to the limited amount of SARS-CoV-2 collected from the swab samples, it was not possible to include biological replicates. The points of same color correspond to technical duplicates of sludge preparation and RNA extraction. At each time point, 2 reactors were sacrificed for analysis (one with unspiked sludge, one with spiked sludge), hence a total of 8 reactors. A value of 0 is indicated when the quantification values were below the detection level. A. SARS-CoV-2 RNA levels in spiked and unspiked sludge, measured by RT-ddPCR targeting the regions N1 + N2 + IP2 (detected in the same channel). B. BCoV RNA levels in spiked sludge, measured by RT-qPCR targeting the N gene. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

For BCoV, the RNA level rapidly decreased over time based on RT-qPCR quantification (Fig. 4B). It was of 59,000 GU/g of total weight at day 0, and no BCoV RNA was detected after 5 days of incubation. Moreover, a good reproducibility was observed among both technical replicates.

BCoV RNA was detected at higher levels than SARS-CoV-2 RNA at day 0, despite a lower spiking level (Fig. 2, Fig. 4). Several hypotheses can explain this observation. BCoV could be more robust than SARS-CoV-2 in the tested conditions. In addition, in the present study, BCoV was produced in controlled conditions whereas SARS-CoV-2 originated from swab samples and could be less well preserved. The detailed results of this quantification assay are provided in the supplementary materials (Table S1).

As a conclusion, the decrease observed over time in all cases, and the total lack of detection of BCoV after 5 days of incubation, support the efficiency of thermophilic AD to degrade SARS-CoV-2 and BCoV. Regarding the methods, the RT-ddPCR analysis successfully increased the sensitivity for SARS-CoV-2 RNA detection during thermophilic AD (Fig. 4), compared to RT-qPCR.

Discussion

To our knowledge, limited information is available on SARS-CoV-2 persistence in WWT sludge matrices (Peccia et al., 2020a, Peccia et al., 2020b; Balboa et al., 2021; D'Aoust et al., 2021; Adelodun et al., 2022), while wastewater matrices have been the subject of considerable interest with the perspective of WBE. Interestingly, the detected levels of SARS-CoV-2 RNA in the raw sludge samples were overall consistent with the number of new COVID-19 cases per day (Table 2, Supplementary Fig. 1). Although the number of fresh sludge samples is limited in our study, these qualitative results are consistent with previous studies and reinforce the notion that WWT sludge could also be used for epidemiological monitoring (Peccia et al., 2020a, Peccia et al., 2020b; Balboa et al., 2021; D'Aoust et al., 2021).

Sludge matrices are challenging as far as RNA extraction is concerned. The stability of coronaviruses in aqueous matrices at 4 °C or 20–25 °C has been widely studied. A synthesis on T90 estimates (time to achieve 90% (one log) reduction for SARS-CoV-2, other coronaviruses and their surrogates, in aqueous matrices, is available in a recently published study (de Oliveira et al., 2021). For SARS-CoV-2 RNA in wastewater, at temperatures comprised between 20 °C and 25 °C, the estimated T90 was of 1.2 days (de Oliveira et al., 2021) or 1.6 days (Bivins et al., 2020) for infectious particles, and of 12.6 days for SARS-CoV-2 RNA (Ahmed et al., 2020a, Ahmed et al., 2020b). At 4 °C, it shifted to 5.5 days for infectious particles (de Oliveira et al., 2021) and 27.8 days for their RNA (Ahmed et al., 2020a, Ahmed et al., 2020b). As a general trend, the decay rate of coronaviruses in wastewater is most of the time higher than in river water (de Oliveira et al., 2021), tap water (Gundy et al., 2008) or reagent grade water (Casanova et al., 2009). A similar observation was reported by Silverman and Boehm (2020), who reviewed the persistence of human coronaviruses and their surrogates in water and wastewater, and assessed the decay rate constants. Between 22 and 25 °C, the decay rate constant was of 0.19 ± 0.06 d−1 in laboratory buffer (n = 3), and of 2.9 ± 0.03 d−1 in sterilized wastewater (n = 4). This is equivalent to a 2 log10 reduction achieved in 24 days and 1.6 days in average, respectively, and to a 3 log10 reduction after 36 and 2.4 days, respectively.

Regarding WWT sludge, a key question is to determine to which extent the sludge matrix has a stabilizing effect on SARS-CoV-2 due to its affinity for solids. Another question is whether the sludge matrix can promote a faster degradation, for instance through biological activity. In the present study, at 4 °C and room temperature, the initial detected levels of viral RNA were low despite the spiking: it was therefore not possible to assess a decay rate. Nevertheless, the high stability of the CT values observed over the monitoring period, on both spiked and naturally contaminated sludge, still suggests the high persistence of SARS-CoV-2 RNA and the surrogate BCoV RNA in fresh WWT sludge at 4 °C or room temperature. This may likely be explained by coronavirus affinity for solids, combined with a limited biological activity at moderate temperatures.

Indeed, the high affinity of SARS-CoV-2 for solids has been highlighted multiple times. Balboa et al. (2021) observed higher SARS-CoV-2 RNA concentrations in the sludge than in the effluent wastewater, in a same studied WWT plant. The authors concluded that SARS-CoV-2 particles were retained in the sludge line. Similar conclusions were drawn by Serra-Compte et al. (2021), by monitoring 16 WWTP, and by ourselves (present study compared to (Lopez Viveros et al., 2021), in revision). In addition, several studies were focused on the detection of SARS-CoV-2 in WWT sludge, as an alternative to wastewater, for epidemiological survey (Peccia et al., 2020a, Peccia et al., 2020b; D'Aoust et al., 2021). However, no definite conclusions can be drawn from the present study regarding SARS-CoV-2 viability in sludge at 4 °C or room temperature. Indeed, the particles' infectivity was not assessed here, and it has been shown for SARS-CoV-2 (Watanabe et al., 2022) and other viruses (Gassilloud et al., 2003; Simonet and Gantzer, 2006) that the genetic material can decay much less faster than infectious particles.

Regarding the thermophilic AD assay, the decay of most SARS-CoV-2 and BCoV RNA was observed within 1 day and 2 days, respectively, in the present study. Such a fast decay was expected since thermophilic AD is generally considered as a sanitizing process. It is also consistent with previous studies on SARS-CoV-2. In Bardi and Oliaee (2021), no SARS-CoV-2 RNA was reported after thermophilic digestion during 150 h (6.25 days) at 50 °C, of food waste and sewage sludge, mixed with urine that contained SARS-CoV-2, originating from patients. By contrast, SARS-CoV-2 RNA was still detected in the same conditions, except in the absence of inoculum (7.3 ± 0.46 × 103 copies/mL). This highlights the role of the inoculum, and thus likely of the biological activity, in SARS-CoV-2 degradation during thermophilic AD. In a more controlled environment, Chin et al. (2020) examined the stability of SARS-CoV-2 at different temperatures and they did not detect any infectious particles after 30 min at 56 °C, in virus transport medium. In wastewater, Bivins et al. (2020) reported a T90 value of 15 min at 50 °C for infectious SARS-CoV-2 particles. For other coronaviruses, the trends are similar. The loss of most of the infectivity is reported within 20 min to a few hours according to the cases, under temperatures exceeding 50 °C. Studies are available for SARS-CoV-1 (Darnell et al., 2004; Rabenau et al., 2005), MERS-CoV (Leclercq et al., 2014) and the transmissible gastroenteritis virus (TGEV, an alphacoronavirus) (Laude, 1981), in various aqueous media. Interestingly (Laude, 1981), also showed that distinct degradation mechanisms are involved above 45 °C, compared to lower temperatures.

In our study, the sole detection of RNA does not inform on the infectivity of SARS-CoV-2 particles and preclude comparing directly our results with the studies evaluating the infectivity of the viral particles. However, our results are still in line with the notion that thermophilic AD is a suitable approach to sanitize fresh WWT sludge in the case of SARS-CoV-2.

Using RT-ddPCR, we resolved the lack of sensitivity problem occurring with RT-qPCR for SARS-CoV-2 detection in sludge during the thermophilic AD assay. It suggests that such a ddPCR-based approach may be useful in future studies, for inhibitor rich matrices or for low concentration targets. Unexpectedly, in another study (D'Aoust et al., 2021), signal inhibition was observed when using RT-ddPCR compared to RT-qPCR on primary clarified sludge, while similar quantification results were obtained on post grit solids. It calls for further efforts to benchmark and standardize methods and procedures for RNA quantification in sludge, a challenging matrix.

Conclusion

Relying on RT-qPCR or RT-ddPCR, we studied, in laboratory reactors, the persistence of SARS-CoV-2 in fresh urban WWT sludge at 4 °C and at room temperature (∼20 °C), as well as during thermophilic anaerobic digestion (50 °C). Levels of RNA from SARS-CoV-2 particles, inactivated or not, were close to the detection limit but highly stable in fresh WWT sludge, over the whole duration of the assays at 4 °C (55 days) and at ambient temperature (∼20 °C, 25 days). By contrast, the RNA levels of BCoV and inactivated SARS-CoV-2 particles rapidly decreased during the thermophilic anaerobic digestion of fresh WWT sludge (50 °C), lasting for 5 days, with final levels that were below or close to the detection limit. This work provides invaluable information regarding the fate of SARS-CoV-2 in fresh WWT sludge, especially since the sludge matrix has been much less studied until now compared to aqueous environments.

In particular, although the infectivity of the particles was not evaluated in the present work, the results strongly support that thermophilic anaerobic digestion is a suitable process for the sanitation of sludge containing SARS-CoV-2, consistent with previous knowledge on thermophilic anaerobic digestion.

Credit author statement

Camille Levesque-Ninio: Conceptualization, Methodology (RNA extraction, RT-qPCR), Supervision, Formal analysis, Writing - original draft. Ariane Bize: Conceptualization, Formal analysis, Visualization, Writing - original draft, Alice Janvier: Investigation (RNA extraction, RT-qPCR), Carlyne Lacroix: Investigation (anaerobic digestion), Florence Le Brizoual: Investigation (RNA extraction, RT-qPCR), Jérôme Barbier: Methodology, Investigation (RT-ddPCR), Formal analysis, Céline Roose Amsaleg: Conceptualization, Writing - review &edit, Sam Azimi: Project administration, Sabrina Guérin-Rechdaoui: Conceptualization, Methodology (anaerobic digestion), Supervision, Writing - review &edit, Funding acquisition, Vincent Rocher: Conceptualization, Supervision, Writing - review &edit, Project administration.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.