Capsid integrity RT-qPCR for the selective detection of intact SARS-CoV-2 in wastewater

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) genomes have been detected in wastewater worldwide. However, the assessment of SARS-CoV-2 infectivity in wastewater has been limited due to the stringent requirements of biosafety level 3. The main objective of this study is to investigate the applicability of capsid integrity RT-qPCR for the selective detection of intact SARS-CoV-2 in wastewater. Three capsid integrity reagents, namely ethidium monoazide (EMA, 0.1–100 μM), propidium monoazide (PMA, 0.1–100 μM), and cis-dichlorodiammineplatinum (CDDP, 0.1–1000 μM), were tested for their effects on different forms (including free genomes, intact and heat-inactivated) of murine hepatitis virus (MHV), which was used as a surrogate for SARS-CoV-2. CDDP at a concentration of 100 μM was identified as the most efficient reagent for the selective detection of infectious MHV by RT-qPCR (CDDP-RT-qPCR). Next, two common virus concentration methods including ultrafiltration (UF) and polyethylene glycol (PEG) precipitation were investigated for their compatibility with capsid integrity RT-qPCR. The UF method was more suitable than the PEG method since it recovered intact MHV (mean ± SD, 38% ± 29%) in wastewater much better than the PEG method did (0.013% ± 0.015%). Finally, CDDP-RT-qPCR was compared with RT-qPCR alone for the detection of SARS-CoV-2 in 16 raw wastewater samples collected in the Greater Tokyo Area. Five samples were positive for SARS-CoV-2 when evaluated by RT-qPCR alone. However, intact SARS-CoV-2 was detected in only three positive samples when determined by CDDP-RT-qPCR. Although CDDP-RT-qPCR was unable to determine the infectivity of SARS-CoV-2 in wastewater, this method could improve the interpretation of positive results of SARS-CoV-2 obtained by RT-qPCR.


H I G H L I G H T S
• Capsid integrity RT-qPCR was optimized to assess intact form of SARS-CoV-2 in wastewater. • CDDP-RT-qPCR was effective for the selective detection of intact enveloped viruses. • UF method was more effective than PEG method for the recovery of enveloped viruses. • Intact SARS-CoV-2 was detected from wastewater. • CDDP-RT-qPCR can provide a better estimate of the presence of SARS-CoV-2 in wastewater.

G R A P H I C A L A B S T R A C T
a b s t r a c t a r t i c l e i n f o

Introduction
The ongoing coronavirus disease 2019  pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a global concern that has spread throughout more than 200 countries and has caused approximately three million confirmed deaths worldwide (WHO, 2021). SARS-CoV-2 is mainly transmitted among people via respiratory droplets and contact routes. However, the presence of high concentrations (up to 10 8 copies/g of feces) of SARS-CoV-2 RNA in the feces of infected people has also been commonly reported Cheung et al., 2020;Lescure et al., 2020;Zhang et al., 2020b). Several studies have successfully isolated viable SARS-CoV-2 from the urine and feces of COVID-19 patients Xiao et al., 2020), thus raising concerns about the possible transmission of SARS-CoV-2 through the fecal-oral route. Recently, it was found that 1.5 and 1.7 days were required for a 1.0 log reduction (T 90 ) of infectious SARS-CoV-2 at room temperature (20°C) in wastewater and tap water, respectively (Bivins et al., 2020). At cool water temperatures, infectious SARS-CoV-2 was able to survive much longer. Indeed, Camilo et al. (2020) reported that the T 90 values of infectious SARS-CoV-2 were 7.7 and 5.5 days and the T 99 (time required for a 2 log reduction) values were 18.7 and 17.5 days in river water and wastewater at 4°C, respectively. This evidence suggests that exposure to wastewater contaminated with SARS-CoV-2 might be associated with certain risks, particularly during the low-temperature season in places where sanitation infrastructure is not available or is insufficient (Bao and Canh, 2021;Pandey et al., 2021).
Capsid integrity RT-qPCR has been recently developed to overcome the limitations of conventional RT-qPCR and thus to discriminate infectious from damaged viruses. In this approach, samples are pretreated with capsid integrity reagents such as monoazide dyes (e.g., ethidium monoazide (EMA), propidium monoazide (PMA) and PMAxx (a new version of PMA)), or more recently metal compounds (e.g., cisdichlorodiammineplatinum (CDDP) and platinum (IV) chloride (PtCl 4 )) prior to RT-qPCR. These reagents can bind to the genomes of damaged viruses with a compromised capsid and subsequently inhibit their detection by RT-qPCR. Therefore, only the genomes of viruses with an intact capsid (potentially infectious viruses) can be detected by capsid integrity RT-qPCR. In recent decades, capsid integrity RT-qPCR has been intensively studied to assess the infectivity of various enteric viruses such as hepatitis A virus, hepatitis E virus, rotaviruses, adenoviruses, noroviruses, and the Aichi virus (AiV) in water and foodstuffs (Coudray-meunier et al., 2013;Fuster et al., 2016;Leifels et al., 2015Leifels et al., , 2021Randazzo et al., 2018;Sangsanont et al., 2014). More recently, this method was also applied to assess the capsid integrity of enteric viruses in tap water Prevost et al., 2016) and to predict the infectivity of noroviruses in bottled water associated with a gastroenteritis outbreak in an epidemiological investigation (Blanco et al., 2017). Therefore, capsid integrity RT-qPCR may be applicable for the selective detection of potentially infectious SARS-CoV-2 in environmental waters. However, most studies thus far have focused on non-enveloped enteric viruses, which have a different structure than that of SARS-CoV-2 (an enveloped virus). Indeed, enveloped viruses such as SARS-CoV-2 are coated with a lipid bilayer membrane that is not present in the structure of non-enveloped viruses. In a recent study, a propidium monoazide compound (PMAxx) combined with RT-qPCR (PMAxx-RT-qPCR) was compared with conventional RT-qPCR and infectivity assay to assess the persistence of different forms of SARS-CoV-2 (free RNA, protected RNA and infectious virus) in wastewater under the impact of water temperature (Wurtzer et al., 2021). However, this study directly applied 100 μM PMAxx without optimizing the PMAxx concentration. More recently, although capsid integrity reagents (monoazide dyes and platinum compounds) were optimized based on free RNA and inactivated SARS-CoV-2 suspended in buffer and urine and stool specimens, the effects of these reagents on infectious or intact SARS-CoV-2 were not investigated (Cuevas-ferrando et al., 2021). In a previous study, the MS2 bacteriophage was found to be inactivated after exposure to a high concentration of capsid integrity reagents (particularly, PMA > 125 μM) (Kim and Ko, 2012). Since little is known about the interaction of capsid integrity reagents on enveloped viruses, capsid integrity RT-qPCR must be developed/optimized for the detection of enveloped viruses, particularly SARS-CoV-2.
In wastewater surveillance strategies for SARS-CoV-2 conducted across the world, virus concentration methods such as ultrafiltration (UF) and polyethylene glycol (PEG) precipitation have been commonly used to recover SARS-CoV-2 RNA before the virus is quantified by RT-qPCR . These virus concentration methods were confirmed to be effective for the recovery of SARS-CoV-2 surrogates (such as murine hepatitis virus (MHV) and Pseudomonas phage φ6) in wastewater samples (Ahmed et al., 2020c;Torii et al., 2021). However, the compatibility of these virus concentration methods with capsid integrity RT-qPCR is concerning because if an intact capsid of the viruses is compromised, capsid integrity reagents could block those viruses from being detected by RT-qPCR and eventually cause false-negative results.
This current study aimed to investigate the applicability of capsid integrity RT-qPCR for the selective detection of intact SARS-CoV-2 in wastewater. Given the stringent requirements for handling SARS-CoV-2 (biosafety level 3), MHV was used as a model virus to optimize capsid integrity RT-qPCR since MHV also belongs to the genus Betacoronavirus and thus has a similar morphology to that of SARS-CoV-2 (Gorbalenya et al., 2020). Additionally, MHV has previously been used as a useful surrogate for investigating the persistence of SARS-CoV-2 in water or evaluating the recovery efficiency of virus concentration methods for SARS-CoV-2 (Ahmed et al., 2020b;Bivins et al., 2020). We tested three capsid integrity reagents (including EMA, PMA, and CDDP) to determine the most effective one for the selective detection of intact MHV by RT-qPCR. Next, the effects of virus concentration methods (including the UF and PEG precipitation methods) on viral structure were evaluated using spiked MHV in raw wastewater. Finally, the optimal concentration method and effective capsid integrity RT-qPCR were applied to assess the presence of SARS-CoV-2 in raw wastewater.

Preparation of intact virus stock and infectivity assay
All the laboratory works were conducted at the Graduate School of Engineering, the University of Tokyo. MHV A59 strain (ATCC VR-764) was propagated using DBT cells. The DBT cells were grown in Eagle's minimum essential medium (MEM, Wako, Japan) supplemented with 5% fetal bovine serum (FBS) in a 75-cm 2 flask. The semi-confluent DBT cells were inoculated with MHV and incubated in MEM with 1% FBS at 37°C (5% CO 2 ) for 3 days. Then, the flask was frozen and thawed once to recover the MHV. The recovered MHV was purified by membrane filtration with a cellulose acetate filter (0.2 μm, DISMIC-25CS, Advantec, Tokyo, Japan) and gel filtration with an Illustra Microspin S-300 HR column (GE Healthcare, Tokyo, Japan) (Kitajima et al., 2010;Sangsanont et al., 2014). The purified MHV was regarded as intact MHV stock. The intact virus stock was aliquoted and maintained at −80°C until use.
The infective MHV was enumerated by the most probable number assay (Meister et al., 2018;Torii et al., 2020) using DBT cells on 96well plates, with four serial dilutions and five replicates. The samples to be quantified were serially diluted by 10-fold using maintenance medium, supplemented with 1% FBS. Each well was inoculated with 150 μL of the diluted samples. After 3 days of incubation at 37°C with 5% CO 2 , the presence of cytopathic effects was monitored by microscopy. The number of positive wells of each dilution was counted and analyzed by an R package (Ferguson and Ihrie, 2019).

Preparation of free viral RNA
Free viral RNA was prepared from purified virus stocks after RNA extraction (as described below). Free viral RNA stock was aliquoted and maintained at −80°C until use.

Capsid integrity treatments
EMA powder (phenanthiridium, 3-amino-8-azido-5-ethyl-6phenyl, bromide; Invitrogen, Carlsbad, USA) was dissolved in Milli-Q water to obtain a stock solution (10 mM). PMA solution (20 mM in H 2 O) was obtained from Biotium, Inc., Fremont, CA, USA. CDDP (powder form, Sigma-Aldrich) was dissolved in DMSO to obtain a concentration of 100 mM. After preparation, the EMA, PMA, and CDDP stock solutions were stored at −20°C in the dark until use. EMA, PMA, or CDDP treatment was conducted according to our previous study (Canh et al., 2019). Briefly, EMA, PMA, or CDDP stock solution (14 μL) was added to the water samples (126 μL) to obtain the desired concentration. EMA or PMA spiked samples were incubated in the dark at 4°C for 30 min and then exposed to a halogen light (650 W, Selecon Pacific, Auckland, New Zealand) for 3 min at a distance of 15 cm from the light source. During the light exposure, the samples were placed on ice water to minimize heating effects. It was confirmed that the light exposure had no damage effects on viral capsid (data not shown). For the CDDP treatment, CDDP spiked samples were incubated at room temperature (20°C) for 30 min. Light exposure was not required for the CDDP treatment. Then, the samples treated with EMA, PMA, or CDDP were subjected to nucleic acid extraction (as described below).

Nucleic acid extraction, reverse transcription, and qPCR detection
Viral RNA was extracted from 140-μL water samples using the QIAamp viral RNA Minikit (Qiagen, Tokyo, Japan) according to the manufacturer's protocol. The extracted viral RNA was subjected to reverse transcription (RT) to synthesize cDNA using the high-capacity cDNA RT kit (Applied Biosystems, Tokyo, Japan).
Real-time PCR (qPCR) was conducted using 20 μL of a reaction mixture that contained cDNA (5 μL), TaqMan Gene Expression Master Mix (10 μL) (Applied Biosystems), forward primer and reverse primer (final concentration, 0.5 μM for MNV and SARS-CoV-2 or 0.3 μM for MHV), TaqMan probe (final concentration, 0.125 μM for MNV and SARS-CoV-2 or 0.4 μM for MHV), and nuclease-free water. Quantification of MHV and MNV was performed using a primer and probe set according to previous studies (Besselsen et al., 2002;Kitajima et al., 2008). SARS-CoV-2 quantification was performing using a primer and probe set (CDC_N1) according to the Centers for Disease Control and Prevention (CDC) in the USA (CDC, 2020). The number of viral genome copies per qPCR reaction was determined from a standard curve using gBlocks or plasmid DNA (Integrated DNA Technologies, Coralville, IA, US) containing the target sequence for each amplification. Standard curves were generated by 10-fold serial dilutions (5 × 10 4 -5 × 10 0 copies/reaction).

Heat treatment
Samples (200 μL each, approximately 10 3 plaque-forming units (PFU)/mL) in 1.5-mL microtubes were incubated at 50, 60, 70, and 80°C for 2 min in a dry block heater (Nissin, Tokyo, Japan), and immediately placed on ice water at the end of the experiment. One sample was maintained at 4°C as a control.
2.6. Virus concentration 2.6.1. Ultrafiltration (UF) The wastewater samples (100 mL) were first centrifuged at 3,500 ×g for 15 min to remove particles. The supernatant was further concentrated using a UF device (Centricon Plus-70; Millipore) with a molecular weight cut-off of 30 kDa according to the manufacturer's protocol to obtain a final volume of 0.4-1.0 mL.

PEG precipitation
PEG precipitation was conducted as described in a previous study . Briefly, sample centrifugation (100 mL) was performed at 3,500 ×g for 15 min to remove particles. The supernatant (80 mL) was supplemented with 8 g of PEG8000 and 4.7 g of NaCl to reach final concentrations of 10% (w/v) and 1.0 M, respectively. The mixture was incubated at 4°C overnight in a shaker. Then, the mixture was centrifuged at 10,000 ×g for 30 min. After carefully discarding the supernatant, the precipitate was resuspended with 10 mM phosphate buffer. The concentrate had a final volume of 1.0 mL.
The most suitable concentration of each capsid integrity reagent was further tested with heat-inactivated MHV to determine the most effective capsid integrity reagent. The MHV sample (10 3 PFU/mL) was heated at 80°C for 2 min. At this heat condition, all MHVs were inactivated according to the infectivity assay (data not shown). Then, heat-inactivated MHV was treated with EMA, PMA, or CDDP at its most effective concentration and subjected to RT-qPCR. All samples were tested in triplicate.
Since sodium deoxycholate (SD) pretreatment has been found to improve the performance of capsid integrity treatments (EMA, PMA, and CDPP) to remove inactivated viruses (e.g., AiV) (Canh et al., 2019), the current study also investigated the effect of SD pretreatment on the performance of optimal capsid integrity treatment on inactivated MHV. For this purpose, heat-inactivated MHV samples (at 80°C for 2 min) were pretreated with 0.1% SD (final concentration) according to a previous study (Canh et al., 2019) and subjected to RT-qPCR with and without optimal capsid integrity treatment (as determined from the previous step). MHV samples without heat treatment (or incubated at 4°C for 2 min) were used as the control. All samples were tested in triplicate.

Screening of virus concentration methods
The UF and PEG precipitation methods were evaluated using raw wastewater to determine the suitable virus concentration method for the application of capsid integrity RT-qPCR. Raw wastewater samples (approximately 500 mL) were collected in June 2020 from wastewater treatment plants (WWTPs) located in the Greater Tokyo Area in Japan. The raw wastewater had the following water quality parameters: temperature, 22.9°C; pH, 7.66; turbidity, 58.2 NTU; electrical conductivity, 556 μs/cm; and UV 254 absorbance, 0.304 cm −1 . The samples were maintained at −20°C until use.
Raw wastewater samples (100 mL) spiked with MHV (approximately 10 8 copies/mL) were concentrated using UF or PEG precipitation (as described above) to obtain the final concentrated samples (0.4-1.0 mL). Then, the final concentrated samples were subjected to RT-qPCR and the most effective capsid integrity RT-qPCR (as determined from the previous section). The experiment was performed in triplicate.
2.9. Detection of SARS-CoV-2 in wastewaters 2.9.1. Sample collection and concentration Sixteen raw wastewater samples were collected from six different WWTPs (A, B, C, D, E, and F) in the Greater Tokyo Area in Japan from January 7th to February 25th, 2021. The water samples were frozen at −20°C until use (1-2 months). The capacity and detailed treatment processes of these WWTPs are shown in Table S1. The general information of COVID-19 epidemic in the Greater Tokyo Area was in the range between 100 and 600 cases per day per 1 million people and more details are shown in the supplemental information.
The most suitable concentration method (determined from the previous section) was used to concentrate the raw wastewater (100 mL) to obtain the final concentrated samples (0.4-1.0 mL). The final concentrated samples were immediately analyzed for SARS-CoV-2 by conventional RT-qPCR and the most effective capsid integrity RT-qPCR (determined from the previous section).

Process controls
The efficiency of the virus concentration process, RNA extraction, and RT-qPCR for the detection of SARS-CoV-2 was evaluated using spiked MHV (whole process control). Briefly, MHV (20 μL, approximately 10 8 copies/mL) was spiked into raw wastewater samples (100 mL) and Milli-Q water (100 mL) as a control. The MHV spiked samples were incubated at room temperature (20°C) for 1.5 h before performing the most effective concentration method (as determined in the previous section). The concentrated samples (140 μL) were then subjected to RNA extraction and RT-qPCR detection. The MHV recovery efficiency (whole process control, E MHV ) was calculated according the following equation: E MHV (%) = (C/C 0 ) × (1/F) × 100, in which C 0 and C are the concentrations of MHV in MilliQ-water sample (without conducting the virus concentration method) and the concentrated sample, respectively. F represents the concentration factor.
The effectiveness of the capsid integrity treatment to eliminate SARS-CoV-2 RNA in water samples was investigated using MNV RNA since MNV has a similar genome structure to SARS-CoV-2 (Positive single-strand RNA). Briefly, free genomes of MNV (1.4 μL, approximately 10 6 copies/mL) were spiked into the concentrated water samples and MilliQ-water samples as the control and was subsequently subjected to the most effective capsid integrity RT-qPCR (as determined from the previous section). The effectiveness of the capsid integrity treatment was identified based on a comparison between the reduction of MNV RNA in the target water samples and the reduction of MNV RNA in MilliQ-water (control).

Effects of different capsid integrity reagents on intact MHV and its free RNA
The effects of different capsid integrity reagents (EMA, PMA, and CDDP) at different concentrations on intact MHV and its free RNA are presented in Fig. 1. As the EMA and PMA concentrations increased from 0.1 to 100 μM, the mean reductions of MHV RNA increased from 0.7 to 4.3 log 10 and from 2.4 to 5.1 log 10 , respectively. CDDP also yielded a mean reduction from 0.5 to 4.2 log 10 as its concentration increased from 0.1 to 1,000 μM. These findings indicate that all assessed capsid integrity reagents could reduce the detection signal of MHV RNA by RT-qPCR and were more effective at higher concentrations.
Adverse effects on intact MHV virions were also evaluated. At concentrations of 0.1, 1.0, 10, and 100 μM, EMA showed mean reductions of intact MHV of 0.3, 1.4, 2.0, and 2.8 log 10 whereas PMA showed mean reductions of 0.4, 0.5, 0.8, and 1.0 log 10 , respectively. CDDP had smaller effects on intact MHV since less than a 0.9 log 10 reduction was observed even at the highest concentration tested (1,000 μM). The reductions of intact MHV by PMA or CDDP were not significantly different from those of the control (in the absence of PMA or CDDP) (p > 0.05, t-test), indicating that these capsid integrity reagents did not affect the detection of intact MHV by RT-qPCR. This result also implies that the intact MHV stock did not greatly contain free genomes of MHV. However, the reduction of MHV was obtained by EMA at concentrations higher than 0.1 μM (p < 0.05). This result suggests that high concentrations of EMA can affect the detection of intact MHV by RT-qPCR.
Previous studies commonly used EMA and PMA at 100 μM and CDDP at 1,000 μM for performing capsid integrity RT-qPCR on non-enveloped enteric viruses without any effects on intact viral form (Canh et al., 2019;Fraisse et al., 2018). These concentrations also showed no inhibitory effects on RT-qPCR amplification. Therefore, it is possible that enveloped viruses such as MHV are more susceptible to capsid integrity reagents compared with non-enveloped enteric viruses. At high capsid integrity reagent concentrations (particularly EMA > 0.1 μM), the structure of MHV can be altered, which allows capsid integrity reagents to access its genomes so that subsequently these genomes are not detected by RT-qPCR. This result is consistent with a previous study indicating that PMA concentrations higher than 125 μM can cause a loss of infectivity for MS2 bacteriophages (Kim and Ko, 2012).
In the current study, the suitable concentration for each capsid integrity reagent was selected based on two criteria to rule out the effects of damage caused to the viral capsid: 1) the reduction of intact MHV did not differ significantly from that of the control and 2) the mean reduction was less than 0.5 log 10 . Based on these criteria, EMA and PMA at a concentration of 0.1 μM and CDDP at a concentration of 100 μM were selected and considered to have a negligible effect on intact MHV. At these concentrations, CDDP (100 μM) was able to result in a 3.4 log 10 reduction of MHV RNA, which was higher than the 0.5 log 10 and 2.4 log 10 reductions obtained by EMA (0.1 μM) and PMA (0.1 μM), respectively.
In a previous study, platinum compounds (e.g., CDDP and PtCl 4 ) was also found to remove SARS-CoV-2 RNA in PBS buffer more effectively than monoazide dyes (e.g., EMA, PEMAX and PMAxx) (Cuevasferrando et al., 2021). Furthermore, PtCl 4 at 2.5 mM was recommended for the application of capsid integrity RT-qPCR because this concentration showed the highest reduction of gamma/heat-inactivated SARS-CoV-2 in fecal suspensions and nasopharyngeal swabs. However, the recommended concentration of PtCl 4 was much higher than the optimal CDDP concentration (100 μM) in the current study. It is possible that complex matrices in the tested samples can interact with PtCl 4 and reduce the effective concentration of PtCl 4 . Therefore, the high concentration of PtCl 4 is needed to ensure the efficient performance of PtCl 4 -RT-qPCR.

Determination of the optimal capsid integrity reagent
To determine the most effective capsid integrity reagent among EMA (0.1 μM), PMA (0.1 μM), and CDDP (100 μM), their performances were tested on heat-inactivated MHV ( Fig. 2A). CDDP was able to remove heat-inactivated MHV (1.7 log 10 ) more effectively than EMA (0.6 log 10 ) and PMA (0.6 log 10 ). In the previous section, CDDP (100 μM) was also more effective than EMA (0.1 μM) and PMA (0.1 μM) at removing MHV RNA (Figs. 1 and 2A). Therefore, CDDP (100 μM) was considered the most effective among the tested capsid integrity reagents for discrimination of the damaged virus. This finding agrees with our previous study indicating that CDDP was more effective than EMA and PMA in discriminating between infectious and inactivated non-enveloped enteric viruses (e.g., AiV) (Canh et al., 2019).
Moreover, the effects of SD surfactant on the performance of CDDP treatment were also investigated to improve the discrimination of inactivated MHV. As shown in Fig  (3.8 log 10 ) than CDDP alone (1.7 log 10 ). However, the combination of SD and CDDP was found to greatly reduce the detection of intact MHV (3.4 log 10 ). It is possible that SD surfactant might damage the MHV capsid since surfactants have been found to dissolve the lipid bilayer membrane of enveloped viruses (e.g., influenza virus) (Kawahara et al., 2018). Therefore, surfactants (particularly SD) should not be used in combination with CDDP for capsid integrity RT-qPCR to assess the infectivity of enveloped viruses. This finding is inconsistent with previous studies indicating that surfactants (e.g., SD and Trition X-100) can enhance the performance of capsid integrity RT-qPCR on non-enveloped enteric viruses with no impact on their capsid or infectivity (Canh et al., 2019). It is possible that non-enveloped viruses might not be susceptible to the effects of surfactant since they do not contain a lipid bilayer membrane in their structure. Based on these results, capsid integrity RT-qPCR with 100 μM CDDP (CDDP-RT-qPCR) was selected as the most effective method and was used for further analyses in this study.

Performance of capsid integrity RT-qPCR to evaluate MHV inactivation by heat treatment
The performance of capsid integrity RT-qPCR using CDDP (CDDP-RT-qPCR) was compared to that of RT-qPCR alone and the infectivity assay to evaluate MHV inactivation after heat treatment at 50, 60, 70, or 80°C (Fig. 3). Although RT-qPCR alone was unable to reduce MHV at any of the temperatures tested, loss of infectivity was observed at 1.6 log 10 at 60°C and more than 2.8 log 10 (below the detection limit) at 70 and 80°C. This result indicates that RT-qPCR alone was unable to discriminate between infectious and inactivated MHV. By contrast, CDDP-RT-qPCR showed a reduction of 1.0 log 10 , 1.0 log 10 , and 1.7 log 10 at 60, 70, and 80°C, respectively, indicating that CDDP-RT-qPCR was more effective in accessing the infectivity of MHV compared with RT-qPCR alone. This result also suggests that CDDP enters heat-inactivated MHV and subsequently blocks the viral genome from being detected by RT-qPCR. However, compared with the infectivity assay, CDDP-RT-qPCR still yielded a lower reduction of MHV, suggesting that CDDP-RT-qPCR did not fully differentiate between infectious and inactivated MHV. This is consistent with previous studies indicating that capsid integrity RT-qPCR overestimated the actual number of infectious viruses when examining non-enveloped enteric viruses after heat, chlorine, and UV treatments Leifels et al., 2015;Prevost et al., 2016;Randazzo et al., 2018). In fact, the loss of capsid integrity was not always correlated with the loss of viral infectivity (Hamza et al., 2011). Therefore, it is possible that viruses can lose their infectivity through alterations on viral capsid structures while still maintaining their capsid integrity to prevent the entry of capsid integrity reagents.
Enveloped viruses such as MHV have an additional lipid bilayer membrane compared with non-enveloped viruses. After inactivation treatment, enveloped viruses might interact with capsid integrity reagents at different efficiencies compared with non-enveloped viruses. In the current study, CDDP-RT-qPCR was able to reduce 3.4 log 10 of MHV RNA (maximum capacity) but only 1.7 log 10 of heat-inactivated MHV even at the highest temperature tested (80°C for 2 min). Compared with our previous studies on enteric viruses (e.g., AiV), CDDP was capable of removing 5.7 log 10 of AiV RNA and 4.8 log 10 of heat-inactivated AiV (80°C for 1 min) (Canh et al., 2019(Canh et al., , 2020. In other words, CDDP yielded a 0.9 log 10 difference in the reduction between inactivated AiV and its free RNA, which was smaller than the 1.7 log 10 difference in the reduction between inactivated MHV and its free RNA. Most likely, the penetration of CDDP into the capsid of a non-enveloped virus (AiV) was more effective than that into in an enveloped virus (MHV) at a similar inactivation condition. However, more studies are necessary to accurately compare between enveloped and non-enveloped viruses to investigate the performance of capsid integrity of RT-qPCR at a similar concentration of virus stock and capsid integrity reagents.

Screening of virus concentration methods for the application of capsid integrity RT-qPCR
Two virus concentration methods (the UF and PEG methods) were investigated to recover MHV in raw wastewater (Fig. 4). The recovery yields were determined by RT-qPCR alone and CDDP-RT-qPCR (the most effective capsid integrity RT-qPCR determined from the previous section) (Fig. 4A). According to RT-qPCR alone, the UF method yielded a MHV recovery of 17.8% ± 7.4%, which was significantly higher than the recovery of 0.034% ± 0.024% obtained by the PEG method. This result agrees with findings from a previous study indicating a greater recovery of φ6 in wastewater samples by the UF method (6.4-35.8%) compared with the PEG method (1.4-3.0%) . When comparing the effectiveness of different concentration methods for quantifying SARS-CoV-2 RNA in wastewater, LaTurner et al. (2021) also found that the UF method yielded a higher concentration of SARS-CoV-2 RNA in wastewater concentrates than did the PEG method. However, the recovery of intact viruses or the effects on viral capsids by these concentration methods were not investigated since the previous studies applied only RT-qPCR or droplet digital RT-PCR to quantify viruses. In the current study, when determined by CDDP-RT-qPCR, the recovery yields of the UF method (38.0% ± 28.7%) were also greater than those obtained by the PEG method (0.013% ± 0.015%) (Fig. 4A). These results suggest that the UF method was able to recover intact MHV more effectively than the PEG method. In a previous study, the UF method was also found to recover infective MHV and Pseudomonas phage φ6 more effectively than the PEG method according to the infectivity assay (Ye et al., 2016). Therefore, it is better to apply the UF method for evaluating virus concentration when investigating the infectivity of enveloped viruses in wastewater.
Furthermore, for the UF method, the recovery efficiency of MHV determined by conventional RT-qPCR (17.8% ± 7.4%) did not differ significantly from the recovery efficiency of MHV determined by CDDP-RT-qPCR (38.0% ± 28.7%) (Fig. 4A). Additionally, the different levels obtained between RT-qPCR alone and CDDP-RT-qPCR in the samples concentrated by the UF method (0.4 log 10 ) were comparable to those in the control (0.4 log 10 ) (MilliQ-water samples without performing the virus concentration methods) (Fig. 4B). These results suggest that the UF method did not significantly affect the structure of MHV. However, in the samples concentrated by the PEG method, the difference (1.2 log 10 ) between RT-qPCR alone and CDDP-RT-qPCR was significantly greater than that of the control (Fig. 4B). This may have occurred because the structure of MHV was partially destroyed by the PEG Solid triangles with asterisk symbols indicate the negative results (>2.8 log 10 ). The reduction limit of PCR-based assays and the IA was 4.6 log 10 and 2.8 log 10 , respectively.
Error bars indicate the standard deviation (n = 3). method. This finding is consistent with a previous study indicating that the PEG method could affect the infectivity of enveloped viruses. Indeed, Ye et al. (2016) found that the infectivity of MHV was decreased by 2 log 10 when MHV was incubated with PEG for 16 h. In this current study, the mixing process of wastewater samples and PEG was conducted for approximately 20 h. Therefore, it is possible that the PEG method in this current study can cause a certain loss of MHV infectivity due to the disruption of PEG on their lipid bilayers (Ye et al., 2016). Besides, in the PEG method, MHV might also be inactivated when incubated in NaCl (1 M) owing to the impact of osmotic lysis (Cordova et al., 2003). Since the UF method displayed a greater recovery of intact MHV and caused less damage to the MHV capsid than the PEG method, it was selected for concentrating SARS-CoV-2 in wastewater in the following steps of this study.

Application of capsid integrity RT-qPCR for the selective detection of SARS-COV-2 in wastewater samples
Sixteen raw wastewater samples collected from the Greater Tokyo Area were concentrated using the UF concentration method and were quantified for SARS-CoV-2 by RT-qPCR alone and CDDP-RT-qPCR (Table 1). The efficacy of the whole detection processes (including the UF concentration method, RNA extraction, and RT-qPCR amplification) was investigated for all wastewater samples using spiked MHV. The recovery efficiencies were mostly greater than 10%, indicating that the entire detection process was efficient for detecting viruses in the tested wastewater samples. Furthermore, the efficiency of CDDP treatment to eliminate free viral RNA present in wastewater was investigated using MNV RNA. The CDDP treatment was able to remove more than 5.7 log 10 of MNV RNA in all concentrated samples, indicating that the CDDP treatment effectively removed free viral genomes present in the concentrated wastewater samples.
Of the 16 raw wastewater samples, SARS-CoV-2 was detected in 5 samples (B1, C1, D1, F1, and F4) with concentrations ranging from 2.3 × 10 3 to 6.6 × 10 4 copies/L when determined by RT-qPCR alone (Table 1). This result is consistent with previous studies conducted in Japan indicating that the concentration of SARS-CoV-2 was <1.1-4.1 × 10 4 copies/L or higher than the detection limit (4.0 × 10 3 -8.2 × 10 4 copies/L) in raw wastewater Hata et al., 2021;Torii et al., 2021). A higher load of SARS-CoV-2 was also reported in raw wastewater collected in other countries, particularly loads of 1.1-3.2 × 10 5 copies/L in Spain (Randazzo et al., 2020), 10 4 -10 6 copies/L in France (Wurtzer et al., 2020b), and 2.6 × 10 3 -2.2 × 10 6 copies/L in the   Netherlands (Medema et al., 2020). This could be due to differences in the prevalence of SARS-CoV-2 in communities among various countries or the different methodologies applied for virus detection. However, the capsid integrity of SARS-CoV-2 was not assessed in these previous studies because conventional RT-qPCR was applied.
In this current study, among the five samples that were positive for SARS-CoV-2 according to the RT-qPCR analysis, only three samples (D1, F1, and F4) remained positive at concentrations ranging from 2.2 × 10 4 to 3.0 × 10 4 copies/L when determined by CDDP-RT-qPCR (Table 1). Therefore, it is possible that SARS-CoV-2 in samples B1 and C1 (which were only positive according to RT-qPCR) had a compromised structure or was completely inactivated. Conversely, SARS-CoV-2 in samples D1, F1, and F4 (positive according to both RT-qPCR and CDDP-RT-qPCR) likely had intact capsids. Although SARS-CoV-2 with an intact capsid in wastewater was detected by CDDP-RT-qPCR, its infectivity was unknown. As discussed in a previous section (Section 3.3), CDDP-RT-qPCR cannot fully distinguish between infectious and inactivated MHV. Indeed, CDDP-RT-qPCR was able to discriminate only 1.7 log 10 of inactivated MHV whereas all MHV (>2.8 log 10 ) lost their infectivity at the high temperature (80°C). Furthermore, the inactivation conditions in natural wastewater might differ from those evaluated in this current study (at high temperatures). In actual conditions, SARS-CoV-2 might lose its infectivity in wastewater at much lower temperatures (such as room temperature). The infectivity of SARS-CoV-2 could also be affected by biological activities or pH levels of wastewater. Therefore, more studies are necessary to accurately determine the infectivity of SARS-CoV-2 in wastewater and to investigate the ability of CDDP-RT-qPCR to discriminate between infectious and inactivated viruses under these inactivation conditions.
Although CDDP-RT-qPCR was unable to determine the infectivity of SARS-CoV-2 in wastewater, this method enhanced the accuracy of determining the probability of the presence of positive SARS-CoV-2 compared with that obtained by RT-qPCR. This finding is in agreement with a previous study suggesting that PMAxx-RT-qPCR could provide a better estimate of the risk of SARS-CoV-2 infection through wastewater compared with conventional RT-qPCR when investigating the persistence of SARS-CoV-2 in wastewater (Wurtzer et al., 2021). Furthermore, in a previous study, quantitative microbial risk assessment (QMRA) was used to estimate the potential risk of SARS-CoV-2 in natural water bodies (Kumar et al., 2021). In this study, viral RNA copies obtained by conventional RT-qPCR were assumed to reflect the maximum infectious concentration of SARS-CoV-2 as PFUs and were used for the risk assessment. Based on the outcome of our current study, it is possible that the previous study greatly overestimated the risk of SARS-CoV-2. Since the ratio of SARS-CoV-2 RNA copies to the infectious concentration (PFUs) is unknown in environmental waters, the use of SARS-CoV-2 RNA copies determined by CDDP-RT-qPCR for QMRA could provide a more accurate assessment of the potential risk of SARS-CoV-2 than the use of SARS-CoV-2 RNA copies determined by conventional RT-qPCR.

Conclusion
Among the capsid integrity reagents (EMA, PMA, and CDDP) tested at different concentrations, 100 μM CDDP was found to be the most suitable for determining viral infectivity since it showed a negligible effect on intact MHV and reduced a higher level of MHV RNA and heatinactivated MHV. Compared with the PEG method to concentrate MHV in wastewater, the UF method yielded a higher recovery of MHV and caused less damage to the structure of MHV. Thus, the UF method appears to be more suitable for the application of capsid integrity RT-qPCR. The UF concentration method followed by CDDP-RT-qPCR was successfully applied for the selective detection of intact SARS-CoV-2 in wastewater collected in the Greater Tokyo Area. Although CDDP-RT-qPCR was unable to determine the infectivity of SARS-CoV-2 in wastewater, this method could improve the interpretation of positive SARS-CoV-2 results obtained by RT-qPCR.