Short-term stability of wastewater samples for storage and shipment in the context of the EU Sewage Sentinel System for SARS-CoV-2

Abstract

In the context of an EU-wide surveillance system for SARS-CoV-2 in wastewater, recommended by the European Commission, this study aims to provide scientific support to the adequacy of transport and storage conditions of samples both in terms of duration and samples temperature. Three laboratories in Slovenia, Cyprus and Estonia investigated the short-term, one-week, isochronous stability of wastewater samples by RT-qPCR based detection of SARS-CoV-2 genes. The results were tested for statistical significance to determine uncertainty of quantification and shelf-life, at testing temperatures of + 20 °C and − 20 °C, relative to reference at + 4 °C. Samples were collected from three urban wastewater treatment plant influents and analysed respectively for SARS-CoV-2 genes N1, N2 (Laboratory 1), N2, E (Laboratory 2) and N3 (Laboratory 3), with various analytical methods. For a period of 7/8 days at + 20 °C, decreasing trends of measured concentrations were observed for all genes resulting in instability according to the statistical analysis, while at − 20 °C the trend of variation was stable only for N1, N2 (Laboratory 1) and N3 (Laboratory 3). Trends for gene E concentrations at − 20 °C (Laboratory 2) could not be tested statistically for stability because of lack of data. Over a period of just 3 days at + 20 °C, the variation was statistically non-significant indicating stability for genes N1, E and N3 for laboratories 1, 2 and 3, respectively. Nonetheless, the outcome of the study presents evidence to support the choice of the selected temperature at which samples shall be preserved during storage before analysis or transport to the laboratory. The conditions (+4 °C, ∼ few days) chosen for EU wastewater surveillance are in accordance with these results, highlighting the importance of stability testing of environmental samples to determine the short-term analytical uncertainty.

Graphical Abstract

1. Introduction

During the COVID-19 pandemic, Wastewater-Based Epidemiology (WBE) has emerged as an important source of additional information to predict and understand the spreading trend of SARS-CoV-2 within the population [1]. The first pan-European Sewage Sentinel System Surveillance was set up by the European Commission to assess the technical and economic feasibility of monitoring the spread of the SARS-CoV-2 virus in wastewater [2]. Since the beginning of the pandemic, several Member States and many countries outside the EU established their own wastewater national surveillance schemes, anticipating the invitation formalised in the European Commission Recommendation of March 2021, which asked Member States (MS) to systematically roll out sewage surveillance and collaborate with the Commission in the building of a so-called EU Sewage Sentinel System for SARS-CoV-2 (EU4S) [3]. Furthermore, the Recommendation declared that:“…putting in place the recommended surveillance system and procedures will also have an added value beyond SARS-CoV-2 surveillance [.]”, since “[.] it will provide an early warning for future possible outbreaks of other pathogens of concern or threats from other pollutants of emerging concern [.]”.

Such outcomes assert the global-scale recognition that WBE can be implemented as an additional support tool to classical epidemiological parameters in monitoring the state of viral outbreaks, in the detection of variants of concern, and in the evaluation of the spread of viral infections within a population [4], [5], [6].

With this outlook, methods’ harmonization and comparability are fundamental to guarantee absence of systematic bias and validity of data obtained from detection of SARS-CoV-2 in wastewater, if those are to be used as a support tool for decision making by water and health authorities. The need for definition of common procedures and mechanisms of collection and handling, processing and analysis of wastewater samples has been widely touched upon in the literature of the past two years [7], [8].

Among the wastewater sample handling procedures for screening for SARS-CoV-2, lack of consistency has been identified regarding the storing temperature of wastewater samples, which can have consequences on detection results of SARS-CoV-2 viral loads [9]. While studies on the influence of temperature on other enveloped viruses in wastewater have identified storage at + 4 °C as appropriate for a short-period of time (1–5 days) [10], SARS-CoV-2 wastewater samples have been kept at + 4 °C, − 20 °C and − 80 °C before analysis [9], [11] and transported at + 4 °C [12], with little information on their stability. The results of a survey on the methodologies applied to wastewater samples for quantification of SARS-CoV-2 by participants of the Pan-European Sewage Sentinel System confirmed variability of storing conditions (+4 °C, <10 °C and −20 °C) that were applied by the respondents [2].

A study by Ahmed et al. (2020) quantified the decay rate of SARS-CoV-2 RNA and of murine hepatitis virus (MHV) in different water matrices, including untreated wastewater, at storing temperatures + 4 °C, + 15 °C, + 25 °C, + 37 °C [10]. The results confirmed 1-log decay in 27.8 ± 4.5 at + 4 °C, while storage at higher temperatures (+25 °C and +37 °C) was discouraged. Markt et al., although using another surrogate (BRSV), partly supported such findings, by comparing composite samples of raw influent wastewater of two different waste water treatment plants (WWTPs) in Austria, stored at + 4 °C for up to 9 days, and − 20 °C for up to 3 days [13]. Besides variability of detected N gene values due to differences in sampled wastewater characteristics, no significant degradation was reported at + 4 °C, while freezing lead to a decreased signal [13]. While these studies are important to support the choice of proper storage temperature, more investigation is required to assess the impacts of varied transport conditions on SARS-CoV-2 concentrations in the sampled wastewater, to harmonize the WBE sampling process. Indeed, the EU-wide WBE exercise involved various sampling locations around Europe, from which wastewater samples were transported to a central analytical facility. Logistics and transport conditions, rather than storage, thus become objective of investigation. As recognized by Gawlik et al. (2012), in case of EU-wide monitoring exercises, it becomes essential to provide information on the short-term transport stability of samples, not only qualitatively, as indicative stability over time, but also quantitatively, as a defined expiry date or shelf life [14].

The isochronous stability exercise (wastewater samples storage at different temperatures for different times) described in this paper thus aims at deepening the understanding of the effect of temperature on SARS-CoV-2 concentration in wastewater during handling and to support the development of homogeneous and comparable operative procedures for national and international wastewater surveillance schemes. The exercise, following Gawlik et al. (2012) study design and statistics, involved three research laboratories, namely the National Institute of Biology, Slovenia, Nireas-International Water Research Centre of the University of Cyprus, Cyprus, and the University of Tartu, Estonia, referred to as Laboratory 1, 2 and 3, respectively, throughout the paper. This study allowed the assessment of short-term stability results relative to the handling of different wastewater samples under specific reproducibility conditions, in an experimental design where concentration methods and targeted genes varied.

2. Material and methods

2.1. Statistics

The stability assessment was done according to an isochronous stability study scheme in collected influent wastewater samples, in which measurements are performed at the same time at the end of the experiment [15]. The testing scheme was designed according to the research developed previously by Gawlik at et al. (2012) and Linsinger et al. (2001).

In an isochronous stability study, the stability and related uncertainty of quantification are based on regression analysis of concentration measurements over time intervals and on significance testing of the regression line [16]. Samples are exposed at varying storage conditions for specified intervals of time (days, weeks or months) and stored back at the chosen reference temperature before being analysed under repeatability conditions (i.e.: one single analytical run).

The stability of the measured gene concentration is evaluated through a linear equation, with the assumption of a linear degradation over time. Thus, the slope and the standard uncertainty of the slope se_b must be considered according to the following equations:

$$\mathit{Slope}=b=\frac{\sum \left(x_i-\overline{x}\right)\left(y_i-\overline{y}\right)}{{\sum \left(x_i-\overline{x}\right)}^2}$$ (1)

$$\mathit{Standard}\mathit{uncertainty}\mathit{ofthe}\mathit{slope}={\mathit{se}}_b=\frac{S_{y*x}}{\sqrt{{\sum \left(x_i-\overline{x}\right)}^2}}$$ (2)

In the reported equations, x_i and y_i are the individual time points and their relative measurement results, $\stackrel{\circledR }{x}$ and $\stackrel{\circledR }{y}$ are the average of all the x_i and y_i over the total number of measurements. It follows that the number of measurements and the time-lapse between measurements affect the uncertainty of the slope se_b (Formula 2), which is inversely proportional to the number of measurements performed and to the distance of time points from the average time.

Isochronous stability data must firstly be analysed for the absence of any evident significant trend of degradation. The significance of degradation trend can be determined using t-test or F-test. Using a t-test, the standard error of the slope can be calculated, with x corresponding to the time values of the measurement points, $\overline{x}$ to the average of all x values and S_y,x to the standard error of the estimate.

$$\mathrm{Standard\; Error\; of\; the\; slope}={\mathrm{u}}_{\mathrm{b}}\frac{S_{y,x}}{\sqrt{\left(x_i-\overline{x}\right)}}$$ (3)

The standard error of the slope, u_b, results to be dependent on the standard error of the estimate (S_y,x) and the number of relative position of the measurement times x.

S_y,x can finally be considered as the standard deviation around the regression line:

$$S_{y,x}=\sqrt{\frac{{\sum \left(y-\hat{y}\right)}^2}{n-2}}$$ (4)

Where y are the measured concentrations, ŷ the concentration from the regression line and n is the total number of measurements. The slope is tested for significance by comparing $\frac{b}{\mathit{ub}}$ to the value of t-statistic with a given confidence level and n-2 degree of freedom. Likewise, the standard error of the slope (Formula 3) can be used to identify, for an acceptable range of uncertainty (i.e., 20%), the shelf-life time or expiry date for a sample. The shelf-life, defined generally as “the length of time a product may be stored without becoming unsuitable for use [17]”, can indeed be interpreted as an expression of the acceptable uncertainty attributable to degradation of SARS-CoV-2 measured gene in wastewater samples. Obviously, if the slope value ( $\frac{b}{\mathit{ub}}$) is significantly different from 0, indicating degradation, the shelf-life is not computed.

2.2. Study design

The study investigates how different handling times and temperatures affect the concentration of SARS-CoV-2 genes in duplicate wastewater samples in comparison to a set of reference conditions, at which SARS-CoV-2 wastewater samples are considered stable.

Reference samples were stored at + 4 °C for 7 days, while testing samples were exposed to the challenging temperature conditions (i.e., +20 °C, −20 °C), for selected time durations (1–8 days). The two samples that remained at + 4 °C for the whole duration of the experiment represent the reference values (0 day at ± 20 °C) for the evaluation of possible decrease of concentration of SARS-CoV-2 genes. At the end of each day (with the exception of day 6 and 7 corresponding to a weekend), two testing samples were transferred to the reference conditions (i.e., +4 °C), while the remaining samples continue the challenge, until the completion of the experiment, as illustrated by the steps in Fig. 1. The duration of the study and the days of transfer from challenging to reference conditions varied slightly among the three laboratories, namely: for the Slovenian National institute of Biology (Laboratory 1), the length of the experiment was 8 days, with transfer days corresponding to 1, 2, 3, 4, 7 and 8; for the University of Cyprus (Laboratory 2), the length of the experiment was 7 days with transfer days corresponding to 1, 2, 3, 4, 5, 7; for the University of Tartu (Laboratory 3), the length of the experiment was 8 days with transfer days corresponding to 1, 2, 3, 4, 6 and 8.

Fig. 1.

Isochronous short-term stability designs executed simultaneously at two different testing temperatures. The analysis of all samples was performed at the end of the testing period, as indicated by the arrow.

Each laboratory applied different analytical procedures, in terms of concentration, extraction and molecular analytical procedures including measured genes, as summarised in Table 1.

Table 1.

Summary of analytical methods used by the participating laboratories.

Laboratory	Analytical method			Measured gene
	Pre-concentration	Extraction	qPCR analysis
LABORATORY 1 National Institute of Biology, Slovenia	Maxwell® RSC Enviro Total Nucleic Acid Kit (Promega, AS1831) and Luciferase control RNA (Promega L4561)	Maxwell® RSC Enviro Total Nucleic Acid Kit (Promega, AS1831)	One-step RT-qPCR on an ABI 7900 fast HT qPCR device (Applied Biosystems)	N1, N2
LABORATORY 2 Nireas-International Water Research Centre of the University of Cyprus, Cyprus	Centricon® Plus-70 (10 kDa MWCO) (Merck Millipore, Burlington, USA)	RNeasy®PowerMicrobiome Kit (Qiagen, Germantown, MD, USA)	RT-qPCR on a CFX96 Real-time system (Bio-Rad)	N2, E
LABORATORY 3 the University of Tartu, Estonia	Centricon® Plus-70 (10 kDa MWCO) (Merck Millipore, Burlington, USA)	RNeasy® Mini Kit (Qiagen, Germantown, MD, USA)	One-step RT-qPCR on a Roche Light Cycler 480 (Roche)	N3

2.3. Wastewater sample collection

Laboratory 1 collected 24-hour flow-dependent composite influent wastewater samples (4 L) from the inlet of the WWTP in Ljubljana on 16.08.2021 using a refrigerated autosampler. WWTP Ljubljana serves 360,000 population equivalent with a maximum flow rate of 157000 m³ day⁻¹. The samples were transferred to National Institute of Biology (Ljubljana, Slovenia), refrigerated in cooling boxes with ice packs, at a temperature of + 4 °C (transit from WWTP to lab < 1 h). After arrival in the laboratory, the samples were divided into the corresponding isochronous test replicate samples, each consisting of 40 mL in falcon tubes and subjected immediately to the incubation scheme depicted in Fig. 1.

Laboratory 2 collected 24-hour composite influent wastewater samples (6 L) from the inlet of the Limassol urban wastewater treatment plant (WWTP) using a refrigerated automated autosampler. The plant serves a total population of 151000 population equivalents and has an average flow of 21480 m³ day⁻¹. The samples were collected on 08.11.2021. Samples were immediately shipped by the WWTP technical staff to the University of Cyprus laboratory facilities (Nicosia, Cyprus), in cooling boxes containing ice packs, at a temperature of + 4 °C. The average transit time from the WWTP to the laboratory was 2 h. Upon their receipt, samples were homogenized and stored at + 4 °C until the start of the experiment. Prior to sample pre-concentration, 50 mL aliquots in sterile falcon tubes (Deltalab) were taken from homogenised influent samples to reach a total of 100 mL volume for each sample, in duplicates.

Laboratory 3 used automated samplers P6 Mini MAXX (Probenahmetechnik GmbH, Germany) to collect 24-hour composite influent wastewater samples at Tartu (∼100000 population equivalents and had an average flow of 21838 m³ day⁻¹) WWTP inlet on 30.08.2021. To obtain the composite sample 96 subsamples (60–80 mL each) were collected after every 15 min. Sample was kept at + 4 °C until transported to the laboratory at the University of Tartu (Tartu, Estonia), with an average transit time of less than 1 h. ISO 5667–10 quality standards were followed during the sampling.

2.4. Analytical methods for SARS-CoV-2 genes determination

Three different in-house available methodologies were applied in the analysis of 24-hour composite influent wastewater samples for determination of different SARS-CoV-2 genes. Samples were collected by the participating institutions according to their established collection protocols (2.2) and subsequently analysed as part of their national surveillance system for SARS-CoV-2. The following paragraphs detail the applied procedures, in terms of pre-concentration, RNA extraction and molecular measurement techniques.

2.4.1. In-house validated method used at laboratory 1 (Slovenian National institute of Biology)

At the laboratory at the National Institute of Biology each wastewater aliquot sample was concentrated and quantified in duplicate using N1 and N2 assays. Standard curves for quantification using SARS-CoV-2 reference RNA were included in same qPCR plate, Luciferase [18] and Pepper mild mottle virus (PMMoV) [19] assays were also tested to control for inhibition as well as for extraction and concentration procedures. The procedure involved both pre-concentration and nucleic acid extraction for 40 mL of wastewater, using a Maxwell® RSC Enviro Total Nucleic Acid Kit (Promega) following the protocol described by the manufacturer and published by Mondal, et al. (2021). The method involves an initial step where the sample is clarified by centrifugation (10 min, 3000 x g, 10 min) and denatured so that the RNA can be bound to a midi column, washed and eluted in 1 mL volume. This volume is immediately subjected to a second step of magnetic bead-based purification in a Maxwell device to a final elution of 70–80 µL volume. Concentrated and purified RNA was stored at − 20 °C and next day quantified using RT-qPCR with CDC USA assays targeting N1 and N2 genes of SARS-CoV-2, as reported in [21], using Fast Virus 1 step master mix (Life technologies) in an Applied Biosystems ABI 7900 fast HT qPCR device with the following cycling conditions: + 50 °C for 5 min for the RT step, + 95 °C for 20 s for initial denaturation followed by 45 cycles of denaturation for 15 s and combined annealing and extension at + 60 °C for 60 s. Luciferase control RNA is externally added to each sample before extraction and analysed with a luciferase specific qPCR assay [22], to monitor for differences in inhibition/extraction. The naturally occurring PMMoV is also analysed (Haramoto et al., 2013) to account for differences in the concentration step, as well as regular negative and positive controls.

2.4.2. In-house validated method used at laboratory 2 (University of Cyprus)

University of Cyprus measured the N2 and E genes of SARS-CoV-2. Prior to sample concentration, large wastewater solids and debris were removed by centrifugation of each sample at 3000 g for 30 min at + 4 °C. The solid pellet was separated from the supernatant that was transferred to Centricon® Plus-70 centrifugal filters (10 kDa) (Merck Millipore) for sample concentration. 50 mL of each sample was centrifuged at 1500 g for 15 min at + 4 °C in the centrifugal filters. Centrifugations of filters were repeated until all sample (100 mL) passed through the centrifugal filter. After the filtration of the whole sample, the filter cups were reversed and centrifuged for another 5 min at 800 g at + 4 °C to collect the sample concentrate. Finally, the concentrate was extracted from the filter cup and was used for RNA extraction. Concentrates were stored at − 20 °C for a time of less than 24 h until nucleic acid extraction. Total RNA was extracted from the filter concentrates using the RNeasy® PowerMicrobiome Kit (Qiagen, Germantown, MD, USA) according to the manufacturer’s protocol [23]. All samples were extracted in duplicate. Elution of nucleic acids was done in 100 µL of nuclease-free water and RNA extracts are stored at − 80 °C until analysis.

A one-step RT-qPCR was used for the detection of SARS-CoV-2 nucleic acids on a CFX96 Real-time system (Bio-Rad). The probe-based reactions used previously published assays targeting the envelope (E) gene and the N2 region of the virus nucleocapsid genetic sequence by Corman et al. (2020b). Primer pairs for E- and N2-specific primers were used in equimolar concentrations (0.1 μM of each primer per reaction). Thermal cycling conditions were used according to Westhaus et al. (2021) for E gene as follows: + 55 °C for 10 min for the RT step, + 95 °C for 1 min for initial denaturation followed by 45 cycles of denaturation for 10 s and combined annealing and extension at + 60 °C for 30 s. The thermal conditions for N2 gene were used according to the Hong Kong University (HKU) protocol published by the World Health Organization (Corman et al., 2020a) for the detection of the novel coronavirus in suspected human cases. More specifically, the RT step took place at + 50 °C for 5 min, followed by + 95 °C for 20 s for initial denaturation, followed by 45 cycles of denaturation for 5 s and extension at + 60 °C for 30 s

The PCR runs were analysed with the Bio-Rad CFX Maestro software (Bio-Rad laboratories).

For quantification, standard curves using in-vitro synthesized single stranded RNA (EURM-019) in buffer [27] were created, containing known concentrations of the genes of interest.

Quality control and quality assurance to determine possible contamination were conducted during each run using positive and negative controls, respectively. Positive controls were included in each reaction using the same in-vitro synthesized single stranded RNA at a concentration of 1 × 10³ copies/µL, which were run in parallel in each RT-qPCR run. Nuclease-free water was used as a negative control, where two aliquots are analysed from the beginning of the assay to control any potential contamination during the RNA extraction, and two more are added during the RT-qPCR reaction to check for contaminations during the nucleic acid amplification. All quantitative assays were performed in duplicate, so the provided copy numbers corresponded the mean of at four values. If not indicated otherwise, a reaction was considered positive if the cycle threshold (CT) was below 40 cycles.

2.4.3. In-house validated method used at laboratory 3 (University of Tartu)

The University of Tartu provided measurements of N3 gene of SARS-CoV-2. After arrival to the lab, samples were preserved by fixation with 10% volume of stop solution (5% phenol: 95% ethanol) and divided into three 50 mL aliquots in 50 mL centrifugation tubes. Larger suspended solids were removed by centrifugation at 4600 g for 30 min (+4 °C). With syringes, the supernatant was carefully transferred and filtered manually through 0.22 µm Sterivex™ GP Sterile Filter Units (Merck KGaA, Germany) to eliminate non-viral particles from the solution. The filtrate (3 ×50 mL) was collected into new 50 mL centrifugation tubes and concentrated in three 50 mL steps using Centricon® Plus-70 Centrifugal Filter Units (molecular weight cut-off 10 kDa) (Merck Millipore, Burlington, USA) at 3500 g for about 30 min in each step. For some of the samples, centrifugation time was longer. Each sample was concentrated down to about 350 µL solution, which was stored at − 20 °C until RNA extraction (up to max 24 h).

RNeasy® Mini Kit (Qiagen, Germantown, MD, USA) was used to extract RNA according to manufacturer’s instructions with two exceptions. First, each sample was adjusted to 600 µL volume by adding MQ water or removing some of the concentrate if it exceeded 600 µL (which was measured and considered in final concentration calculations), and after adding 600 µL RTL lysis buffer, 800 µL of 96% ethanol solution was added. Secondly, elution was carried out in two steps using 50 µL RNase Free Water (8000 g for 1 min). OneStep™ PCR Inhibitor Removal Kit (Zymo Research, USA) was used for additional removal of PCR inhibiting compounds. RNA extracts were analysed by using quantitative one-step reverse transcription real-time PCR (RT-qPCR). The reaction mixture in a total volume of 10 µL included 3 µL of RNA extract, 1 × One-step Probe CoV Mix, 1x One-step SOLIScript® CoV Mix (Solis BioDyne, Estonia), 200 nM of previously published nucleocapsid N3 primers and fluorescently labelled probe (Microsynth, Switzerland). To reach the final volume, reaction volume was adjusted with molecular grade water (Solis BioDyne, Tartu, Estonia). To quantify SARS-CoV-2 RNA in the samples, a calibration standard curve was constructed for each run using serial dilution series of calibrated EURM-019 single-stranded RNA (EC Joint Research Centre). All the reactions were performed in six replicates. Ultrapure molecular grade water was used as a negative control. RT-qPCR reactions were performed at + 55 °C for 30 min, followed by + 95 °C for 10 min and 45 cycles of + 95 °C for 10 s and + 55 °C for 30 s on Roche LighCycler 480 (Roche Life Sciences, Switzerland).

3. Results

The results obtained in the isochronous stability study are presented in Figs. 2–6, for each of the participating laboratories. The graphs show, for the whole duration of the experiment (0–8 days), the trend of variation of the SARS-CoV-2 gene concentrations (Y), normalised by the gene concentration of the samples kept for 0 days at ± 20 °C (i.e.: reference samples kept at +4 °C for the entire duration of the experiment). The study of the trend of the ratio (Y/Y0) allows to eliminate variations in units of measurements in the expressions of the results among different laboratories.

Fig. 2.

Stability curves of SARS-CoV-2 N1 gene (Laboratory 1) at + 20 °C and − 20 °C.

Fig. 6.

Stability curves of SARS-CoV-2 N3 gene (Laboratory 3) at + 20 °C and − 20 °C.

In case of decreasing trends calculated for the whole duration of the experiments, the statistical test has been carried out for a shorter duration of 3 days. Three days duration reflects the common handling times experienced in the SARS-CoV-2 EU-wide wastewater monitoring exercise conducted by the European Commission. Results are summarised in paragraph 3.4 and Table 2 reports a summary of the statistical analysis, while detailed results of detected concentrations of respective genes are presented in Table A1 (Supplementary Material), along with more extensive statistical calculations.

Table 2.

Results of statistical analysis of Short-term isochronous stability test.

		+ 20 °C				-20 °C
Lab.	SARS-CoV-2 Measured gene	Slope‡ 0	ustd [%]	Slope‡ 0	ustd [%]	Slope‡ 0	ustd [%]	Slope‡ 0	ustd [%]
		7 days		3 days		7 days		3 days
1	N1	YES		NO	26.5	NO	31.2	-	-
	N2	YES		YES		NO	35.2	-	-
2	N2	YES		YES		NAa		YES
	E	YES		NO	63.6	NAa		NAa
3	N3	YES		NO	21.5	NO	57.2	-

^aNA not available

3.1. Results by laboratory 1

For the storing at + 20 °C, all the samples resulted in quantification cycle Cq values within 31.5–37.9, with measured concentrations of 1.3 × 10⁵-1.3 × 10⁷ Cp/L, for N1, and for N2 within 31.5–38.1 Cq with measured concentrations of 2.3 × 10⁵-2.2 × 10⁷ Cp/L. From the 4th day on, the signal (Cq values) increased above the limit of quantification (LOQ) of the method, reaching even undetermined results at samples that were held longer times at + 20 °C, for both genes. For the storing at − 20 °C, all the samples resulted in Cq values within 33.1–34.7, with measured concentrations 1.5–3.9 × 10⁶ Cp/L, for N1, and for N2 within 32.49–34.7 Cq, with concentrations in the range of 2.7 × 10⁶ - 1.1 × 10⁷ Cp/L. All values were within the LOQ of the method and inspection of the Cq values and resulting Cp/L did not suggest any evident degradation of the sample, even at day 7.

The described trends in form of ratio of concentration over the average of the initial concentration (Y/Y0) are reported in Fig. 2 and Fig. 3, for N1 and N2 respectively. A declining trend is observed at + 20 °C in graphs for both gene N1 and gene N2, thus not allowing to calculate shelf-life and uncertainty for the whole duration of the experiment. When considering a duration of 3 days, the slope of Y/Y0 trend for N1 results to be not statistically different from 0 with a calculated uncertainty of quantification of 26.5% and a shelf life_{(uncertainty 20%)} of 2 days, while for N2 it was statistically different. At − 20 °C, the stable trend of the regression lines allowed the calculation of the uncertainty for 8 days storage (31.1% and 35.2%), for N1 and N2 respectively and the shelf-life_{(20% uncertainty)} which resulted in 5 days for both.

Fig. 3.

Stability curves of SARS-CoV-2 N2 gene (Laboratory 1) at + 20 °C and − 20 °C.

3.2. Results by laboratory 2

According to the analysis carried out at the Laboratory 2, at + 20 °C reported concentrations of genes N2 and E were in between 1.81 × 10²-5.6 × 10³ Cp/mL, and 5.15 × 10¹-3.12 × 10³ Cp/mL, respectively, which correspond to a range of Cq values of 33.11–39.76 for N2 and 34.9–39.21 for E gene. At − 20 °C, the minimum detected concentrations were 3.18 × 10² (N2) and 9.8 Cp/mL (E), while the maximum concentrations were 3.40 × 10³ Cp/mL (N2) and 3.59 × 10³ Cp/mL (E) which corresponded to a range of Cq values of 35.19–38.14 for N2 and 36.16–38.70 for E gene. Measurements for day 4 onwards and day 3 onwards for samples stored at − 20 °C were under limit of detection (LOD) for genes N2 and E, respectively.

At + 20 °C, for E gene (Fig. 4) Y/Y0 trend was statistically different from 0 over 7 days duration, while it became equal to 0 only considering a timespan of 3 days, with a calculated uncertainty of 63.3% and a shelf-life _{(20% uncertainty)} of 1 day. In addition, high coefficients of variation resulted from the analysis of replicates. The large variation between replicates of the same sample observed in the initial E gene measurements may be attributed to the varying sensitivity of the E detection assay in real influent wastewater samples. The N2 gene (Fig. 5) concentration decreases during the first two days of storage of samples at + 20 °C. After the estimation of the slope and its standard error (Table 2), it was shown that the signal of the N2 gene marker declines for both 7 and 3 days.

Fig. 4.

Stability curves of SARS-CoV-2 E gene (Laboratory 2) at + 20 °C and − 20 °C.

Fig. 5.

Stability curves of SARS-CoV-2 N2 gene (Laboratory 2) at + 20 °C and − 20 °C.

Storage of samples at − 20 °C was shown to lead to large and rapid decreases in E gene marker signal after 2 days, after which data are not available.

In the same manner, there is a decrease in the N2 signal which continues for the first 3 days of storage of the wastewater samples.

3.3. Results by laboratory 3

Reported concentration of gene N3 spanned between 1.56 × 10³-1.28 × 10⁴ Cp/mL at + 20 °C and 25.82 × 10²-5.43 × 10³ Cp/mL at − 20 °C with corresponding Cq values within 33.2–39.6.

At + 20 °C, for N3, according to the carried out statistical analysis, the trend of Y/Y0 was statistically different from 0 over a storage period of 8 days, while it resulted to be equal to 0 considering a 3 day shelf-life frame with a calculated uncertainty of 21.5% (Fig. 6).

At − 20 °C, on the contrary, the stability result is affected by the last data point reported for the measurement at day 8. Although not a statistical outlier, the higher concentration reported biases the regression towards stability in the case of 8 days study length, with an uncertainty of 57.2% and shelf-life_{(uncertainty 20%)} of 3 days, while, when only three days are considered, decline of measured signal is reported.

3.4. Summary of results

Table 2 summarises the results of the stability test for all laboratories, with no uncertainty reported in case of the slope being significantly different from 0 in the statistical t-test, as described in 2.1.

Summarising, for a period of 7 days at + 20 °C, Y/Y0 decreased for all genes resulting in instability according to the statistical analysis, while at − 20 °C the ratio resulted stable only for N1, N2 (Laboratory 1) and N3 (Laboratory 3).

Over a period of 3 days at + 20 °C the variations of Y/Y0 were non-statistically significant indicating stability for genes N1, E and N3 for the three laboratories respectively. Trends for gene E concentrations at − 20 °C could not be tested statistically for stability because of lack of data.

4. Discussion

The scenario regarding available data on stability of SARS-CoV-2 in wastewater samples reported in previous literature is complex, because of a limited number of studies, different experimental approaches used and genes measured. Even in this study, each participating laboratory has utilized different wastewater processing approaches and investigated different gene markers, depending on their SARS-CoV-2 wastewater surveillance method in place. Moreover, the application of different pre-concentration, RNA extraction and qPCR approaches likely contributes to the heterogeneity of experimental results, and to the variability in detection of SARS-CoV-2 genetic markers. Regarding diversity in measured genes, no previous studies were found to focus on the stability of SARS-CoV-2 detection using the E gene in urban wastewater samples. E gene was shown to sometimes have a lower response among SARS-CoV-2 genetic fragments, compared to N1 and N2 genetic markers (Ahmed et al., 2022). In a similar manner, the N1 assay was shown to have a higher sensitivity of measurement among N1, N2 and E assays in the study by Acosta et al. (2021). According to Markt et al. (2021), the stability of the N1 gene was shown after storage at + 4 °C for 9 days (using PEG for precipitation, TRIzol® for concentration and for extraction); this is in accordance with the results of Laboratory 1 for which detection of N1 was stable for 8 days (using Promega Maxwell® RSC Enviro Total Nucleic Acid Kit for concentration and for extraction). While Markt et al. (2021) found that freezing was not beneficial, as N1 concentrations quickly declined after storage at − 18 °C for 3 days, similarly to Laboratory 2 results for N2 gene, results from Laboratory 1 confirmed stability up to 8 days even when stored at − 20 °C. The same study indicates that freezing-thawing was shown to lead to a decrease in signal for N1, an aspect which needs to be taken into consideration when handling wastewater samples (Markt et al., 2021). The differences observed upon freezing, storing at − 20 °C between Markt et al. and Laboratory 1 in this study, can likely be linked to the different methodologies used for concentration.

Similarly to Markt et al. (2021), the study by Islam et al. (2022) has investigated the effect of storage of urban wastewater influents from 6 different WWTPs for 7 and 14 days at + 4 °C, − 20 °C and − 80 °C (using PEG for concentration, Qiagen RNeasy® PowerMicrobiome Kit for extraction). The results have shown a decrease in N gene signal after storage of the samples at − 20 °C for 14 days compared to the concentrations measured at + 4 °C.

Fernandez-Cassi et al. (2021) also reported a notable reduction in SARS-CoV-2 signals in raw influent wastewater stored at − 20 °C for one month (using Centricon® -Plus 70 for concentration, Qiagen RNA Viral Mini Kit for extraction). These findings suggest that the possible freeze-thaw process that takes place after storage at − 20 °C, may have a significant impact on influent wastewater samples. These findings are in contrast with Hokajärvi et al. (2021), where no genetic marker signal decay was reported after 58 days of sample storage, at − 20 °C and − 80 °C of the genetic markers N2 and E (using Centricon® -Plus 70 for concentration, Chemagic Viral300 DNA/RNA kit for extraction). This was the only study where the observation length was so extended.

In another instance [31], samples were stored at + 4 °C, room temperature or + 37 °C for 24 h before investigation of N1 and N2 genetic marker concentrations (using PEG for concentration, Qiagen QIAmp RNA Viral Mini Kit for extraction). There was no statistically significant difference in signal among samples stored at any of the tested temperatures, which could be explained by the fact that an increase in temperature can confer enzyme activity impairment in urban wastewater, which would otherwise have biodegraded SARS-CoV-2 genetic markers. Indeed storage at + 20 °C, as resulted from the isochronous study described here, for more than 3 days is not advisable as the signal seems to rapidly degrade, independently of the gene.

As it is clear from results of the laboratories involved in this study and from reviewed literature, the stability of SARS-CoV-2 in wastewater seems to be dependent, among others variables, on the concentration method chosen and the detected gene. This is not unexpected as the different methods make use of different viral features for their concentration, and the viral degradation should affect them differently. Additional variability might be caused by the wastewater matrix of collected samples, which itself can be influenced by sewage system infrastructural and operational conditions. Nonetheless, the evaluation of all the possible impact factors is not the purpose of this study, which instead aims at identifying proper storage and transport conditions that allow, under controlled uncertainty, the execution of wastewater-based surveillance programmes at a national and international scale.

5. Conclusions

This isochronous stability study aimed at identifying proper storage and transport conditions that allow, under controlled uncertainty, the execution of wastewater-based surveillance programmes on national and international scale. The chosen reference temperature of + 4 °C seems to be the more cautious and practical suggestion for handling wastewater samples for SARS-CoV-2 detection, provided that the analysis could be performed within a limited time frame of few days from collection. Indeed, as per recommendation of the EC and the recently WHO published guidance on “Environmental surveillance for SARS-COV-2 to complement public health surveillance” [32] storing at + 4 °C during the whole duration of samples transport is the recommended measure.

This isochronous stability study should serve as first step in understanding and establishing context for proper storage and transport conditions of urban wastewater samples. The statistical evaluation depends on the interplay of sampling procedure, analytical method and detected genes, and therefore it is advisable that, while WBE becomes more implemented as a support tool in monitoring of SARS-CoV-2 and other pathogens, isochronous stability testing under various respective laboratory environments is carried out to define proper transport and storing conditions of wastewater samples. To guarantee statistical strength, the isochronous stability studies should be structured according to the actual storage and transport conditions of each context and should be conducted under repeatability conditions.

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.

Editor: V. Victor

Appendix

See in Table A1.

Table A1.

Results of stability and statistical analysis for the laboratories at testing temperatures for the respective measured genes.

		[Y/Y0]
	Testing Temperature	T = +20 °C		T = - 20 °C			Testing Temperature	T = +20 °C		T = - 20 °C			Testing Temperature	T = +20 °C	T = - 20 °C
	Days/Gene	N1	N2	N1	N2		Days /Gene	N2	E	N2	E		Days /Gene	N3	N3
Laboratory 1	0	1.171	1.019	0.831	0.787	Laboratory 2	0	1.043	0.392	0.978	1.995	Laboratory 3	0	0.817	0.757
	0	0.829	0.981	1.169	1.133		0	0.957	1.608	1.022	0.005		0	1.183	1.243
	1			0.480	0.461		1	0.560	0.920	0.786	0.736		1	0.698	0.685
	1	1.304	0.818	0.608	0.419		1	0.806	0.811	0.691	0.328		1	1.314	0.546
	2	1.116	0.542	0.636	0.614		2	0.248	0.517	0.264			2	1.420	0.274
	2	0.859	0.539	1.109	0.466		2	0.334	0.487	0.240			2	1.173	0.395
	3	0.767	0.621	1.293	0.646		3	0.376	0.030	0.096			3	1.079	0.277
	3	1.320	0.622	0.565	0.427		3	0.309	0.782	0.104			3	0.959	0.278
	4	0.558	0.176	0.874	0.359		4	0.129	0.027				4	0.274	0.145
	4	0.339	0.114	0.492	0.341		4	0.169	0.042				4	0.319	0.133
	7	0.102	0.152	0.723	0.302		5	0.054					6	0.240	0.172
	7	0.163	0.036	0.600	0.298		5	0.046					6	0.173	0.175
	8	0.122	0.062	1.262	0.947		7	0.034					8	0.187	0.765
	8	0.074	0.029	1.215	0.509		7	0.051					8	0.238	0.691
	Statistics						Statistics						Statistics
	df	11	11	12	12		df	12	8	6	2		df	12
	b	-0.145	-0.116	0.007	-0.024		b	-0.132	-0.239	-0.319			b	-0.136
	ub	0.031	0.015	0.033	0.024		ub	0.020	0.082	0.027			ub	0.031
	b/ub	-4.759	-7.750	0.206	-1.006		b/ub	-6.505	-2.930	-11.898			b/ub	-4.390
	Slope ‡ 0	YES	YES	NO	NO		Slope ‡ 0	YES	YES	YES			Slope ‡ 0	YES
	Uncertainty (3d)			12%	13%		Uncertainty (3d)						Uncertainty (3d)
	Shelf life (U=20%)			5.1	4.5		Shelf life (U=20%)						Shelf life (U=20%)