To evaluate and compare the reductions of human viruses and F-specific coliphages in a full-scale wastewater treatment plant based on the quantitative PCR (qPCR) and plate count assays.
To evaluate and compare the reductions of human viruses and F-specific coliphages in a full-scale wastewater treatment plant based on the quantitative PCR (qPCR) and plate count assays.
A total of 24 water samples were collected from four locations at the plant, and the relative abundance of human viruses and F-RNA phage genogroups were determined by qPCR. Of the 10 types of viruses tested, enteric adenoviruses were the most prevalent in both influent and effluent wastewater samples. Of the different treatment steps, the activated sludge process was most effective in reducing the microbial loads. Viruses and F-RNA phages showed variable reduction; among them, GI and GIII F-RNA phages showed the lowest and the highest reduction, respectively.
Ten types of viruses were present in wastewater that is discharged into public water bodies after treatment. The variability in reduction for the different virus types demonstrates that selection of adequate viral indicators is important for evaluating the efficacy of wastewater treatment and ensuring the water safety.
Our comprehensive analyses of the occurrence and reduction of viruses and indicators can contribute to the future establishment of appropriate viral indicators to evaluate the efficacy of wastewater treatment.
Waterborne viral diseases are of major concern in both developing and developed countries. Wastewater is often laden with viruses, and wastewater treatment plays a crucial role in mitigating viral pollution of aquatic environments (Bofill-Mas et al. 2000; da Silva et al. 2007; Haramoto et al. 2008; Rodríguez-Díaz et al. 2009). Traditionally, bacterial indicators such as Escherichia coli, total coliforms (TCs), and faecal coliforms have been used to assess microbiological safety of treated wastewater. However, bacterial indicators are not appropriate predictors of the fate of viral pathogens during wastewater treatment because of their different susceptibility to the treatment (Baggi et al. 2001).
F-specific coliphages (F-phages) have been proposed as indicators for the presence of enteric viruses in water environments because their morphological characteristics, such as size and structure, are similar to those of most enteric viruses (IAWPRC 1991). F-phages are divided into F-DNA and F-RNA phages based on their genomic features, and F-RNA phages are further divided into four genogroups, genogroup I (GI) to genogroup IV (GIV). A previous study found that GII and GIII F-RNA phages are generally derived from humans, whereas GI and GIV are derived from other animals such as pigs, cows and chickens (Furuse 1987). Recent studies also suggested that the persistence of F-RNA phages in the environment or during wastewater treatment process varies depending on their genogroup (Cole et al. 2003; Nappier et al. 2006; Haramoto et al. 2012), implying a need for further comprehensive analyses of the occurrence and fate of different F-RNA coliphage genogroups in wastewater.
Several types of human viruses are known to be present in the water environments (Rodríguez-Díaz et al. 2009; Symonds et al. 2009). To establish suitable viral indicators, it is essential to better understand the behaviour of different types of viruses. However, only a few studies have investigated the occurrence and persistence of each virus and F-RNA phage genogroup in wastewater (Muniesa et al. 2009; Haramoto et al. 2012).
The present study aimed to characterize the fate of F-RNA phage genogroups during wastewater treatment as indicators of human viruses. Wastewater samples were collected at a full-scale wastewater treatment plant (WWTP) and the occurrence and reduction of human viruses and F-RNA phage genogroups were determined by quantitative PCR (qPCR) and compared with those of traditional microbial indicators determined by plate counting assays.
Wastewater samples were collected monthly on dry days with antecedent dry weather for several days between October 2007 and March 2008 from a full-scale WWTP located in an urbanized area in Japan. This WWTP has a capability to process 450 000 m3 of wastewater per day and employs a conventional activated sludge process followed by chlorination and sand filtration with hydraulic retention time (HRT) of 9·3 h. The average biochemical oxygen demand of the influent and effluent wastewater samples was 220 and 1 mg l−1, respectively.
A total of 24 water samples were collected from four locations in the treatment train, that is, influent (IN), after activated sludge (AS), after chlorination (CL) and effluent after sand filtration (EF) (Fig. 1). For the CL and EF samples, sodium thiosulphate was added at a final concentration of 50 mg l−1 immediately after sampling to quench residual chlorine. Water samples were collected in 2-l sterile plastic bottles, stored on ice and immediately transported to the laboratory, where they were processed within 12 h from the time of sampling.
Water samples were concentrated by an adsorption–elution method, using an electronegative membrane (cat. no. HAWP-090-00; Millipore, Tokyo, Japan) (Katayama et al. 2002). Briefly, 2·5 mol l−1 of MgCl2 was added to the samples to obtain a final concentration of 25 mmol l−1. The samples (100 ml of IN and 1000 ml each of AS, CL or EF) were then passed through the electronegative membrane filter attached to a glass holder (Advantec, Tokyo, Japan). Magnesium ions were then removed by passing 200 ml of 0·5 mmol l−1 H2SO4 (pH 3·0) through the filter, and viruses were eluted with 10 ml of 1·0 mmol l−1 NaOH (pH 10·8). The eluate was recovered in a tube containing 50 μl of 100 mmol l−1 H2SO4 (pH 1·0) and 100 μl of 100× Tris-EDTA buffer (pH 8·0) for neutralization. Our previous study showed that the recoveries of poliovirus type 1 spiked into influent and effluent wastewater using this method were 23 ± 19% (n = 20) and 65 ± 28% (n = 24), respectively (Katayama et al. 2008).
Eluates were further concentrated using a Centriprep YM-50 (Millipore) filter unit to obtain a final volume of 700 μl, according to the manufacturer's protocol. The concentrated samples were stored at −80°C until further analysis.
Viral RNA and DNA were extracted using the QIAamp Viral RNA Mini Kit (Qiagen, Hilden, Germany) and the QIAamp DNA Mini Kit (Qiagen), respectively, according to the manufacturer's protocol. The nucleic acid extracts were stored at −80°C until further analysis.
The extracted viral RNA was subjected to RT using a High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA). Briefly, 10 μl of RT reaction mixture was prepared by mixing 5 μl of extracted RNA, 1 μl of 10× reverse transcription buffer, 0·4 μl of 25× dNTPs, 1 μl of 10× random hexamers, 25 units of MultiScribe™ (Applied Biosystems) reverse transcriptase and 10 units of RNase inhibitor. The mixture was incubated at 25°C for 10 min, at 37°C for 120 min and at 85°C for 5 min to synthesize cDNA.
TaqMan-based qPCR assays for GI, GII and GIV noroviruses (NoVs), sapoviruses (SaVs), astroviruses (AstVs), enteroviruses (EVs), enteric adenoviruses (E-AdVs), JC polyomaviruses (JCPyVs), BK polyomaviruses (BKPyVs), TT viruses (TTVs) and GI, GII, GIII and GIV F-RNA phages were performed with an ABI PRISM 7500 Sequence Detection System (Applied Biosystems). The sequences of primers and probes were obtained from previous studies, as shown in Table 1. Reaction mixtures (25 μl) consisted of 2× TaqMan® gene expression master mix (Applied Biosystems), 400 nmol l−1 each of forward and reverse primer, 100 nmol l−1 of TaqMan probe and 5 μl of DNA or cDNA template. The reaction mixtures were subjected to thermal cycling as follows: 95°C for 15 min to activate the DNA polymerase, followed by 50 cycles of denaturation at 94°C for 15 s and annealing and extension at specific temperatures for each template (Table 1) for 1 min. Fluorescence reading were collected and analysed using Sequence Detection Software ver. 1.3 (Applied Biosystems). The genome copy numbers of each virus or phage were determined based on the standard curve prepared with 10-fold serial dilutions of plasmid DNA or oligo-DNA, containing each virus gene sequence to be amplified, at a concentration of 104 to 100 copies per reaction.
|Target||Function||Sequence (5′ → 3′)a||Annealing temperature (°C)||References|
|GI NoVs||Forward primer||CGYTGGATGCGNTTYCATGA||56||Kageyama et al. (2003)|
|GII NoVs||Forward primer||CARGARBCNATGTTYAGRTGGATGAG||56||Kageyama et al. (2003)|
|GIV NoVs||Forward primer||TTTGAGTCYATGTACAAGTGGATGC||56||Trujillo et al. (2006)|
|SaVs||Forward primer||GAYCASGCTCTCGCYACCTAC||62||Oka et al. (2006)|
|AstVs||Forward primer||CCGAGTAGGATCGAGGGT||55||Le Cann et al. (2004)|
|EVs||Forward primer||CCTCCGGCCCCTGAATG||60||Shieh et al. (1995)|
|Probe||FAM-CCGACTACTTTGGGTGTCCGTGTTTC-TAMRA||Katayama et al. (2002)|
|E-AdVs||Forward primer||AACTTTCTCTCTTAATAGACGCC||55||Ko et al. (2005)|
|JCPyVs||Forward primer||AGAGTGTTGGGATCCTGTGTTTT||60||Dumonceaux et al. (2008)|
|BKPyVs||Forward primer||ATCCAAACCAAGGGCTCTTTTC||60||Dumonceaux et al. (2008)|
|TTVs||Forward primer||CGGGTGCCGDAGGTGAGTTTACAC||62||Tokita et al. (2002)|
|GI F-RNA phages||Forward primer||GTCCTGCTCRACTTCCTGT||58||Wolf et al. (2008)|
|GII F-RNA phages||Forward primer||TCTATGTATGGATCGCACTCG||58||Wolf et al. (2008)|
|GIII F-RNA phages||Forward primer||GYGGTGCYACAACRACGAAT||58||Wolf et al. (2008)|
|GIV F-RNA phages||Forward primer||GACWGGTCGGTACAAAGTKG||58||Wolf et al. (2008)|
|MNV||Forward primer||CGGTGAAGTGCTTCTGAGGTT||56||Kitajima et al. (2008)|
Murine norovirus (MNV, S7-PP3 strain), kindly provided by Dr Y. Tohya (Nihon University, Kanagawa, Japan) and propagated in RAW 264·7 (ATCC TIB-71) cells (American Type Culture Collection, Manassas, VA, USA), was used as a sample process control to determine the efficiency of extraction–RT-qPCR because the extraction and/or RT-qPCR may be inhibited by substances in the concentrate (Hata et al. 2011). Prior to RNA extraction, 1·4 μl of MNV stock was spiked in 140 μl of concentrated wastewater sample and Milli-Q water as the control. MNV-RNA was co-extracted with other indigenous viral RNA from the water samples, and the MNV-RNA yield was determined by RT-qPCR. The relative MNV% recovery efficiency (R) was calculated as follows:
where C represents the observed MNV-cDNA copy numbers per qPCR tube in a wastewater sample, and C0 represents copy numbers in the control; C0 was determined to be 9·3 × 105 copies per qPCR tube. R values of <100 (%) indicate losses of viral RNA during extraction and/or the occurrence of RT-PCR inhibition. R values were determined for each sample, but they were not used to adjust the concentration of indigenous viruses.
Escherichia coli and TCs in wastewater samples were quantified using a single-agar-layer method with Chromocult® coliform agar (Merck, Darmstadt, Germany). After incubation at 37°C for 24 h, blue colonies were counted as E. coli, while the total blue and red colonies were counted as TCs.
F-phages were quantified by plaque assay using the host strain Salmonella enterica serovar Typhimurium WG49, in accordance with the ISO standard 10705-1 (Anon 1995).
To characterize the persistence of each type of microbes in the wastewater treatments, the log10 reduction in each or whole treatment process was compared using Tukey's multiple comparison test, assuming that the observations were independent and the data from the different groups have a normal distribution with equal variation across observations. Based on this comparison, the different microbes were classified into four groups (labelled 1–4): group 1 microbes showed significantly lower log10 reductions than those of any microbes in the other groups (P < 0·01). Group 2 microbes showed significantly lower log10 reductions than those of group 4 microbes (P < 0·01), and group 3 microbes were intermediate between those of group 2 and group 4.
The log10 reduction in activated sludge or whole treatment process was also compared by likelihood ratio test after fitting the reductions to beta distribution, as detailed in previous studies (Teunis et al. 2009; Schijven et al. 2011; Schmidt and Emelko 2011).
To monitor the overall efficiency of RNA extraction-RT-qPCR, a known amount of MNV was spiked to the concentrated water samples as a process control. The recovery efficiency of MNV from all samples was >30% (Table 2), showing good performance of the analytical scheme.
|Sample||Number of samples||Geometric mean (Range)|
Viruses and F-RNA phage genogroups were determined by RT-qPCR, and microbial indicators (E. coli, TCs and F-phages) were assayed by plate count methods. All viruses tested were consistently detected in the IN samples during all months, with a mean concentration of >104 copies l−1 (Fig. 2). E-AdVs were the most abundant, with a geometric mean of 2·2 × 106 copies l−1, followed by SaVs (5·1 × 105 copies l−1), GII NoVs (2·9 × 105 copies l−1) and JCPyVs (2·1 × 105 copies l−1) (Fig. 2). All viruses except for TTVs were also detected in the EF samples, with E-AdVs again being the most abundant (4·4 × 103 copies l−1).
Of the four F-RNA phage genogroups, three (GI, GII and GIII) were detected in both IN and EF samples, whereas GIV was not detected in any of the samples. GII F-RNA phages were detected at the highest concentration in the IN sample (1·0 × 105 copies l−1), while GI F-RNA phages showed the highest concentration in the EF sample (5·2 × 103 copies l−1). The microbial indicators TCs, E. coli and F-phages were also detected in all samples. F-phages were detected in 50 times higher numbers than GII F-RNA phages (Fig. 2).
Table 3 summarizes log10 reductions of viruses and microbial indicators for each of the different treatment processes separately as well as for the entire treatment. For the complete process, overall reduction of TTVs could not be evaluated as they were not detected in the EF samples. However, the reduction can be estimated to be >1·63 log10 assuming a detection limit of five genome copies per reaction, which amounts to 3·3 × 102 copies l−1 in the EF sample. Based on Tukey's multiple comparison test of the log10 reductions for the whole treatment process, viruses and phages could be classified into four groups (Table 3). For the caliciviruses, GI and GIV NoVs were classified into group 2, while GII NoVs and SaVs were classified into group 3 and 4, respectively. Similarly, GI, GII and GIII F-RNA phages were classified into group 1, 3 and 4, respectively. Reductions were also compared by likelihood ratio test after fitted to beta distribution (Tables S1 and S2), but this analysis failed to yield a similar classification. Nevertheless, this analysis verified that GI F-RNA phages, GIV NoVs and GI NoVs showed significantly lower reductions than any other microbes and that SaVs and GIII F-RNA phages showed significantly lower reductions than some other microbes.
|Viruses and microbial indicators||Groupa||Mean log10 reduction ± standard deviation|
|Whole process||Activated sludge||Chlorination||Sand filtration|
|GI F-RNA phages||1||0·49 ± 0·29||0·83 ± 0·25||0·09 ± 0·54||−0·43 ± 0·45|
|GIV NoVs||2||1·29 ± 0·34||1·40 ± 0·67||0·15 ± 0·37||−0·26 ± 0·58|
|GI NoVs||1·58 ± 0·22||1·65 ± 0·26||0·38 ± 0·72||−0·45 ± 0·53|
|GII F-RNA phages||3||2·04 ± 0·50||2·50 ± 0·35||−0·19 ± 0·46||−0·12 ± 0·26|
|BKPyVs||2·18 ± 0·51||2·29 ± 0·73||0·09 ± 0·45||0·08 ± 0·34|
|EVs||2·29 ± 0·69||2·20 ± 0·12||0·12 ± 0·42||0·49 ± 0·03|
|GII NoVs||2·35 ± 0·40||1·81 ± 0·53||0·49 ± 0·44||0·05 ± 0·22|
|AstVs||2·63 ± 0·08||2·42 ± 0·35||0·13 ± 0·22||N.Ab.|
|E-AdVs||4||2·69 ± 0·65||1·88 ± 0·43||0·11 ± 0·25||0·71 ± 0·57|
|F-phages||3·08 ± 0·49||2·67 ± 0·42||0·24 ± 0·20||0·16 ± 0·44|
|JCPyVs||3·19 ± 0·63||2·99 ± 0·67||−0·33 ± 0·34||0·44 ± 0·42|
|SaVs||3·31 ± 0·55||3·28 ± 1·13||−0·04 ± 1·29||0·07 ± 0·65|
|GIII F-RNA phages||3·39 ± 0·55||2·86 ± 0·25||0·14 ± 0·22||0·21 ± 0·44|
|TCs||4·07 ± 0·29||2·46 ± 0·28||0·47 ± 0·23||1·13 ± 0·34|
|Escherichia coli||4·74 ± 0·25||2·72 ± 0·31||0·42 ± 0·27||1·60 ± 0·36|
|TTVs||Unclassified||N.A.||1·60 ± 0·37||−0·15 ± 0·21||N.A.|
The activated sludge process was more effective in reducing microbial loads, rather than chlorination and sand filtration processes, although the microbes could not be classified into the different groups by Tukey's multiple comparison test or likelihood ratio test after fitting to a beta distribution. Still, GI F-RNA phages, GIV NoVs, TTVs and GI NoVs showed significantly lower reductions than microbes efficiently reduced such as SaVs, JCPyVs and GIII F-RNA phages (Tables S3 and S4). The log10 reductions in bacterial indicators, E. coli and TCs, by sand filtration were >1, whereas those of viruses and phages were much lower (Fig. 2, Table 3).
In this study, we investigated the occurrence and reduction of viruses and microbial indicators including F-RNA phage genogroups at a full-scale WWTP in Japan.
Many types of viruses were detected at high concentration in the wastewater flowing into the plant (IN sample), indicating that these are prevalent in the study area. All viruses but TTVs were also found in the fully treated water (EF sample), suggesting that the WWTP effluent is a source of viral contamination to receiving water bodies. In both IN and EF samples, E-AdVs were most abundant. Previously, AdVs have been proposed as an indicator of human faecal contamination because of their high prevalence in water environments (Albinana-Gimenez et al. 2009; McQuaig et al. 2009), which is in accordance with our findings. EVs, JCPyVs and BKPyVs have also been proposed as viral indicators (Bofill-Mas et al. 2000; Hot et al. 2003) but were less abundant in our samples.
Human caliciviruses, such as NoVs and SaVs, are causative agents of gastroenteritis and are frequently found in both clinical and environmental samples, especially during the winter season (Phan et al. 2006; Haramoto et al. 2008; Nordgren et al. 2009; La Rosa et al. 2010; Kitajima et al. 2011; Sano et al. 2011; Sima et al. 2011). Our results also revealed high prevalence of caliciviruses in winter. Interestingly, the levels of GIV NoVs, which is generally less common in both patient and environmental samples (Iritani et al. 2002; La Rosa et al. 2008, 2010; Kitajima et al. 2011; Sima et al. 2011), were comparable with those of the other NoV genogroups. This illustrates that analyses of wastewater may be an appropriate approach for assessing the prevalence of viruses that are less frequently observed in epidemiological surveys. Although SaVs have not been identified as often as NoVs from gastroenteritis patients (Haramoto et al. 2008), they were detected at higher concentration than NoVs from the IN samples, indicating that SaVs may be more prevalent in water environments than previously appreciated. Also, SaV concentrations in the wastewater samples analysed in the present study were higher than those reported in wastewater samples collected from another Japanese WWTP during 2005–2006 (Haramoto et al. 2008), suggesting that the abundance of SaVs varies greatly from year to year, as was also reported based on epidemiological studies (Pang et al. 2009).
The recovery of MNV spiked into IN samples was higher than those spiked into treated samples (AS, CL and EF samples), indicating that the microbial reductions assessed by RT-qPCR may be overestimated. The AS process was the most effective process for microbial reduction. Of the caliciviruses, GIV NoVs was most persistent, whereas SaVs were reduced more efficiently. Similarly, of the F-RNA phages, GI F-RNA phages were the least reduced during treatment. Overall, these results suggest that treatment efficiency is highly dependent on the geno- or serotypes of viruses.
It is well known that qPCR may overestimate the numbers of intact virus, especially after disinfection which inactivates viruses and allows their nucleic acids to remain (Shin and Sobsey 2008; Pecson et al. 2009). Here, two different methods were employed to quantify F-phages. The plaque assay was used to quantify viable F-phages, and RT-qPCR was used to quantify genome copies of each F-RNA phage genogroup regardless of their viability. Interestingly, the reduction of viable F-phages as determined by both techniques was comparable. This suggests that reduction in the activated sludge process was mainly due to physical removal of viral particles, rather than the inactivation of viruses.
In Japan, the concentration of TCs in the effluent wastewater is regulated and may not exceed 3000 CFU ml−1. All EF samples tested in this study were below the regulatory level. This underscores that bacterial indicators alone fail to indicate water safety against enteric viruses as many types of viruses and phages were detected after sand filtration. While infectivity assays were not conducted, the presence of viable F-phages strongly suggests the presence of viable viruses in these samples (Tree et al. 2003). Thus, it is important to consider the microbial safety of recreational areas or water sources where effluent wastewater is discharged. The introduction of additional treatments, such as UV irradiation (Lee et al. 2008) and ozonation (Ishida et al. 2008), could be effective in inactivating viruses.
F-phages are potential indicators of viral contamination and treatment efficiency (IAWPRC 1991). However, their characters, such as origin, prevalence and persistence in the water environment and survivability during water treatment, are highly dependent on their sero- or genogroups (Schaper et al. 2002; Nappier et al. 2006; Muniesa et al. 2009; Haramoto et al. 2012). In IN samples, the concentration of F-phages determined by the plaque assay was 24 times higher than that determined by RT-qPCR even though RT-PCR assay for viruses generally shows higher count than infectivity assays because of the presence of nonviable particles. With respect to F-phages, relatively higher portion of viable particles might be present in the samples. Besides, presence of F-phages that cannot be detected by our RT-qPCR assay, such as F-DNA phages and/or other F-RNA phages, might also attribute to the difference (Haramoto et al. 2009, 2012). In the future study, plaque assay with and without RNase treatment should be employed to determine the fraction of F-RNA and F-DNA phages.
Of the F-RNA phages, GIV F-RNA phages were not detected in any of the samples tested. Previous studies also documented low incidence of GIV F-RNA phages in human faeces and water environments (Ogorzaly and Gantzer 2006; Love et al. 2008; Wolf et al. 2008; Ogorzaly et al. 2009; Haramoto et al. 2012). GII and GIII F-RNA phages are generally excreted by humans, while GI and GIV are excreted by animals (Furuse 1987). In the present study, GI F-RNA phages were frequently detected in the wastewater samples. Presence of GI F-RNA phages generally indicates contamination by animal faeces; however, Cole et al. (2003) showed that human faeces also contain GI F-RNA phages. As the WWTP is located in an urbanized area, the source of GI F-RNA phages is likely to be human and/or domestic pets.
MS2 and Qβ are the representative strains of GI and GIII F-RNA phages, respectively, and it has been well documented that MS2 is much more persistent under environmental stresses and water treatment than Qβ (Schaper et al. 2002; Nappier et al. 2006; Muniesa et al. 2009). In addition, the reduction of GI F-RNA phages during treatment determined in the present study was the lowest among the tested microbes, while that of GIII F-RNA phage was the highest, although the reduction was comparable with that of some other viruses (Table 3). These observations indicate that GI F-RNA phage and GIII F-RNA phage may be useful as viral indicators to represent the minimal and maximal reduction of viruses during the wastewater treatments, respectively. The most abundant F-RNA phage genogroups were GII and GI in the IN and EF samples, respectively, indicating that the dominant genogroup had shifted during wastewater treatment, which is in accordance with a previous study (Muniesa et al. 2009; Haramoto et al. 2012). This is probably due to the low reduction of GI F-RNA phages. Reduction of F-phages determined by plaque assay was also similar to those of several other viruses, indicating their potential utility in predicting the reduction of the viral load during wastewater treatment. An advantage with F-phages is that they can be detected by plaque assay, which only detects viable phages. A combination of F-RNA phage genogroup-specific RT-qPCR assays and plaque assay is useful for predicting behaviours of viruses during wastewater treatment.
In this study, we quantified various types of human viruses and F-RNA phage genogroups at a full-scale WWTP in Japan. Our results demonstrate that at least 10 types of human viruses, including GIV NoVs, were prevalent among humans in the study area and discharged into public water body from WWTP. Of the different treatment steps, the AS process was the most effective in reducing viruses and phages. The variability in reductions among viruses and F-RNA phage genogroups after wastewater treatments substantiates that selection of an adequate viral indicator is important for ensuring water safety and to estimating the efficacy of wastewater treatment. Our study provides valuable information contributing to reveal the prevalence of viruses in wastewater, their fate during treatment and the future establishment of appropriate viral indicators in treated wastewater.
This work was supported by Core Research for Evolutional Science and Technology (CREST).