Julie Jean, Département des Aliments et de Nutrition, Institut des Nutraceutiques et Aliments Fonctionnels (INAF), Université Laval, Québec, QC, Canada, G1K 7P4. E-mail: Julie.Jean@fsaa.ulaval.ca
Aims: To evaluate the persistence of human norovirus (NoV) in different types of water at various temperatures using conventional and TaqMan real-time reverse transcription-PCR (RT-PCR).
Methods and Results: Water from different sources was spiked with NoV and incubated at different temperatures over a 3-month period. NoV viral RNA was amplified by one-step TaqMan real-time RT-PCR and by conventional two-step RT-PCR. NoV persisted in mineral and tap water for over 2 months at all tested temperatures but disappeared after 100 days. At 4 and −20°C, viral degradation was slower than that at 25°C. In river water and effluent from primary sewage treatment, a slight reduction in viral load was observed after 1 month at 4°C. This is the first demonstration of medium-to-long-term survival of human NoVs in different types of water using TaqMan real-time detection.
Conclusions: NoV genome may persist for long periods of time in different types of water. Quantitative TaqMan real-time RT-PCR is a sensitive system that allows accurate evaluation of the persistence of human NoVs in different water samples.
Significance and Impact of the study: Our study is one of the few to demonstrate the ability of NoV to survive for a long time in water.
To our knowledge, no evaluation of the persistence of human NoVs in various water samples has ever been reported in real-time RT-PCR. The objective of this study was therefore to evaluate the persistence of human NoV-GII genome in water from different sources using TaqMan real-time RT-PCR and conventional RT-PCR.
Materials and methods
A stool sample was obtained during a gastroenteritis outbreak in March 2005 and tested positive for NoV-GII-4, by enzyme-linked immunosorbent assay (Dakocytomation, Ely, UK) was kindly provided by the Newfoundland and Labrador Public Health Laboratory. The identity of the norovirus in the sample has been confirmed by conventional RT-PCR as described below and by sequencing (GenBank accession no. DQ367878).
Effluent from the anaerobic digestion stage of sewage treatment without UV treatment (St Nicolas, Quebec, Canada), mineral water (local supermarket), tap water (Quebec City, Canada) and river water (St Lawrence River, Canada) were spiked with GII-NoV. Briefly, 0·3 ml of effluent or water was spiked with 20 μl [corresponding to 100 RT-PCR units (RT-PCRU) after end-point titration] of stool sample diluted 10-fold in diethylpyrocarbonate (DEPC)–water. The spiked tap water and mineral water samples were then incubated at 4, 25 and −20°C for 20, 40, 60, 80 and 100 days.
River and sewage water samples were incubated for 10, 20 and 30 days at 4 and 25°C. The experiments were repeated in duplicate.
Nucleic acid extraction
Nucleic acid extraction from the incubated samples and unspiked water samples (negative controls) was based on the Boom method (Boom et al. 1990) using the Basic Extraction kit (for tap water or mineral water) or the Magnetic Extraction kit (for river water or sewage treatment effluent) according to the manufacturer’s instructions (BioMerieux, Boxtel, the Netherlands). Samples were mixed with 1 ml of lysis buffer and 50 μl of silica beads and incubated for 10 min. After several washes, RNA was eluted from the silica with 50 μl of elution buffer and stored at −80°C or 10-fold serially diluted for immediate amplification.
Primers targeting the conserved ORF1 (open reading frame 1) region, which includes the GLSPG motif of the RNA-dependent RNA polymerase (primer SR46: 5-TGGAATTCCATCGCCCACTGG; Ando et al. 1995) and the highly conserved ORF1–ORF2 junction region (primer GOG2R: 5-TCGACGCCATCTTCATTCACA; Kageyama et al. 2003), were used for NoV detection. These primers amplify a 587-bp segment of the NoV genome.
Reverse transcription-PCR was run in two steps. The reverse transcriptase reaction was performed at 42°C for 2 h in a 20-μl volume with 5 μl of viral RNA extract, 10 pmol of primer GOG2R (Applied Biosystems, Foster City, CA, USA), 0·5 μg of poly dT(18) primer (Amersham Biosciences, Buckinghamshire, UK), 1 μl of 25 mmol l−1dNTPs (deoxynucleoside triphosphates) (Promega, Madison, WI, USA) and 40 units of M-MuLV reverse transcriptase (New England Biolabs, Pickering, ON, Canada).
For the PCR, 7 μl of cDNA (containing 0·5 mmol l−1 dNTPs; Promega), two units of Taq DNA polymerase enzyme (New England Biolabs), 1x ThermoPol Buffer (New England Biolabs) and 10 pmol of primers SR46 and GOG2R were used. The negative control was prepared with sterile (DEPC)–water and the positive control with NoV-GII-4. A Mastercycler gradient (Eppendorf, Hamburg, Germany) was used and the conditions were as follows: initial denaturation at 95°C for 5 min, 30 cycles at 95°C for 1 min, 55°C for 1 min and 72°C for 1 min, followed by a final extension at 72°C for 10 min. The control PCR was amplified with Ando’s primers for GII (SR46, SR33) and generated a 123-bp fragment (Ando et al. 1995). The amplified products were then analysed by 1% agarose gel electrophoresis and visualized with ethidium bromide.
TaqMan real-time RT-PCR
Real-time RT-PCR was performed using a TaqMan one-step RT-PCR Master Mix Reagents Kit (Applied Biosystems), following the procedure described by the manufacturer. The reaction took place in a final volume of 20 μl and contained 5 μl of viral RNA extract, 800 nmol l−1 of each primer, namely JJ2F (5′-CAAGAGTCAATGTTTAGGTGGATGAG, Jothikumar et al. 2005) and GOG2R, 150 nmol l−1 of labelled probe RING2 (6FAM-TGGGAGGGCGATCGCAATCT-TAMRA; Kageyama et al. 2003), 0·5 μl RT/RNase Inhibitor Mix and 10 μl TaqMan® One-Step RT-PCR Master Mix (Applied Biosystems). A 7500 Real-Time PCR (Applied Biosystems) was used. The reverse transcriptase step was run at 48°C for 30 min, followed by initial denaturation at 95°C for 10 min to activate the TaqMan polymerase. PCR consisted of 40 cycles of melting at 95°C for 1 min, annealing and extension at 56°C for 1 min. A DNA segment of 98 bp in length was thus amplified. Fluorescence of FAM liberated from the probe by TaqMan was measured to determine the amplification threshold cycle (Ct), which was the first cycle at which fluorescent emission was 10-fold higher than the standard deviation of the mean baseline emission. Negative controls were performed with 5 μl of sterile DEPC–water. A serial 10-fold dilution of extracted viral RNA (40 ng μl−1) was used to establish the standard curve. The threshold cycles were set manually after PCR was completed to generate the standard curve; the unknowns were then calculated based on the standard curve. To determine the RNA concentration, the absorbance of the reaction mixture was measured at 260 nm using an ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA). The viral reduction percentage was determined by the estimation of recovery virus in the spiked sample using the method described by Fuhrman et al. (2005). In brief, this paper describes the method used to estimate the percentage of viral recovery or reduction in different water samples.
Sensitivity of the one-step TaqMan real-time vs two-step conventional RT-PCR
Using a 10-fold serial dilution of the RNA extract (40 ng μl−1, from the diluted stool sample), conventional RT-PCR showed an amplification detection limit of 10−3, which was designated as one RT-PCR amplifiable unit or RT-PCRU (Fig. 1). Based on previous results, detection of 0·0001 RT-PCRU was obtained by TaqMan real-time RT-PCR, with Ct mean values of 38·16 and 37·60 (Fig. 2b). The sensitivity of TaqMan PCR is thus 10 000 times greater than that of conventional RT-PCR. An excellent correlation between the target template concentration and the Ct was observed (Fig. 2a,b).
Evaluation of NoV genome persistence in water samples by conventional RT-PCR
Figure 3 shows the survival of NoVs in different water samples at 4, 25 and −20°C, as revealed by gel electrophoresis of RT-PCR-amplified RNA. No significant difference in band intensity was observed between 40 and 60 days (lanes 5–7 and lanes 10 and 11). However, a significant decrease in band intensity was observed at 25°C after 80 days in tap water (lane 11), probably because of reduced viral concentration. After 3 months, no signal was obtained in tap water samples.
Persistence of NoV genome in mineral water shows similar results to those obtained with tap water (Fig. 3, lanes 2–4 and lanes 8 and 9). Specific amplification was obtained until 80 days of incubation. No amplification was obtained after 100-day incubations (data not shown). In the case of sewage treatment effluent and river water, specific amplicons were obtained by RT-PCR after 30 days of incubation at 4°C (Fig. 4, lanes 5 and 6).
Evaluation of NoV genome persistence in water samples by TaqMan real-time RT-PCR
At 4°C in mineral water, a 30% reduction in NoV concentration was observed after 40 days of incubation (Fig. 5a, Table 1). A reduction between 76% and 86% was obtained after 60 and 80 days and 100% after 100 days. At 25°C, the reduction was 63% after 40 days (Fig. 5b, Table 1) and 83%, 95% and 99·98% at 60, 80 and 100 days, respectively. At −20°C, reduction was about 12% after 40 days (Fig. 5c, Table 1), while no NoV was detected after 80 and 100 days.
Table 1. Viral reduction in mineral and tap water samples after different incubation periods at various temperatures, as determined by real time reverse transcription-PCR.
Temperature samples (°C)
*Each experiment was analysed in duplicate.
10·09 ± 1·6
30·3 ± 1·8
76·09 ± 0·7
86·58 ± 1·9
100 ± 0·2
46·22 ± 2·0
86·7 ± 1·9
97·08 ± 2·2
99·91 ± 2·1
99·98 ± 0·8
9·2 ± 1·8
63·68 ± 1·7
83·04 ± 2·1
95·08 ± 2·0
99·98 ± 0·2
31·69 ± 2·1
67·23 ± 1·8
71·21 ± 1·9
84·23 ± 1·6
99·89 ± 0·9
8·3 ± 2·5
12 ± 2·1
45·85 ± 1·9
98·5 ± 0·6
100 ± 0·1
15·23 ± 2·2
25 ± 3·9
41·15 ± 2·0
100 ± 0·5
100 ± 0·2
In tap water, 86%, 97% and 99·9% reductions in NoV concentration were observed after 40, 60 and 80 days, respectively, at 4°C (Fig. 5d, Table 1). At 25°C, a somewhat similar trend was observed, with reductions ranging from 67% to 84% over 40–80 days (Fig. 5e, Table 1), while reductions of 25% and 41% were obtained after 40 and 60 days, respectively, and reduction below detectable levels by 80 days at −20°C (Fig. 5f, Table 1).One hundred days was sufficient to reduce the fluorescence signal of NoV RNA to negative control levels at all tested temperatures.
The persistence of NoV genome in river water at 4°C was similar to that obtained in effluent from sewage treatment. After 30 days of incubation, viral concentration was reduced to 15% for river water and 21% for sewage-treated water of the initial value (Fig. 6a,c, Table 2). At 25°C, reductions were 67% in river water (Fig. 6b) and 49% in sewage treatment effluent (Fig 6d, Table 2).
Table 2. Viral reduction in river and sewage-treated water samples after different incubation periods at various temperatures, as determined by real time reverse transcription-PCR
Our data show that human NoV genome can persist in all tested conditions (4, 25, and −20°C) for up to 2 months. However, complete degradation of NoV genome was observed after 100 days. Viral RNA was not detected at this point, indicating that the persistence of NoVs may be from rare viral particles that degrade only after a long time.
Results from real-time amplification were correlated with those from conventional RT-PCR assays, although there were sensitivity differences between the assays. Our quantitative assay use an RNA standard curve, to estimate the viral charge in water and shows similar sensitivity with other TaqMan assays that uses serial dilution of GII Plasmid DNA (Kageyama et al. 2003; Jothikumar et al. 2005; Loisy et al. 2005; Hoehne and Schreier 2006). The detection limit obtained with our two-step RT-PCR corresponded to the 10−3 dilution of the RNA extract from the stool sample (diluted to 10%), while the TaqMan real-time system was 10 000 times more sensitive.
To date, only one study on persistence of enteric viruses HAV (hepatitis A virus), FCV (feline calicivirus) and NoVs in marinated mussel tissues was described using real-time PCR and cell culture system (Hewiitt and Greening 2004). This study has shown a 1·7 log reduction by culture methods for hepatitis A and no reduction in NoV or HAV RT-PCR titre over 4 weeks. The NoV real-time method used in their study was 2 log more sensitive than conventional two-step RT-PCR.
Our results for viral persistence using human NoVs show significant differences from those already reported using surrogate viruses or other enteric viruses. Several papers have shown that surrogate viruses are not very persistent in the environment. For example, Allwood et al. (2005) have shown that FCV and MS2 (male-specific bacteriophage) survive in tap water at 4°C for 28 days, while FCV was undetectable after 21 days at 25°C and after 7 days at 37°C.
The persistence of human NoV genome in mineral water can be explained by the presence of nutrients or minerals that have a possible protective effect. Hurst et al. (1989) have already demonstrated the protection of enteric viruses in water, although the effects of these compounds on NoV genome persistence were not investigated in their study.
We used magnetic silica beads for RNA extraction from sewage treatment effluent or river water because this method has been proven to increase the detectable amount of viral RNA (Rutjes et al. 2005).
Noroviruses added to river water and sewage treatment effluent were detectable at only slightly reduced levels after 1 month by RT-PCR and by TaqMan real-time PCR. Our data confirm the persistence of NoV genome in sewage treatment effluent and its potential risk to human health. Ueki et al. (2005) demonstrated the geographically associated circulation of NoV between humans and cultivated oysters via waterways in sewage, river water and treated wastewater.
The temperature is a significant factor affecting pathogen inactivation and our data suggest that the viral reduction or degradation was lower at 4 or −20°C compared to 25°C. These results may be explaining the predominance of NoV during cooler seasons (Haramoto et al. 2006).
The persistence of other enteric viruses in water has been described. Enteric poliovirus and hepatitis A virus persisted for 1 year in mineral water at 4°C and 120 days at room temperature (Biziagos et al. 1988). Hurst et al. (1989) examined the long-term survival of coxsackievirus B3, echovirus 7 and poliovirus in freshwater and showed 6·5–7 log10 units of viral inactivation over 8 weeks at 22°C, 4–5 log10 units over 12 weeks at 1°C and 0·4–0·8 log10 units over 12 weeks at −20°C. The authors concluded that nutrient concentration, water turbidity and suspended solids affect the survival of viruses in water.
A significant scale of epidemiological evidence has shown that waterborne routes of exposure are critical in the transmission of human NoVs and/or other enteric viruses (Schaub and Oshiro 2000). Our results showed that low levels of NoV RNA were detectable over 2 months of incubation at different temperatures in four tested water types. We have also demonstrated that the use of a NoV RNA standard curve combined with magnetic extraction and the monoplex one-step TaqMan real-time PCR assay can maximize the quantification of low viral loads in prolonged incubations. Rzezuka and Cook (2004) illustrated the persistence of human enteric viruses in the environment and in food and showed the lack of information on NoV persistence. Furthermore, the study of NoV genome persistence remains a challenge as the possibility of detecting RNA may be derived from either nonviable or viable viruses. Recently, Straub et al. (2007) have developed an in vitro cell culture infectivity assay for human NoV. This study opens new perspectives in survival studies for human NoV. These authors have also demonstrated a clear correlation between the presence of infectious virus by cytopathic effect and detectable human noroviral RNA by RT-PCR. The real-time RT-PCR protocol for GII NoVs presented in this present study can be applied to survival studies in environmental and food samples in which NoVs are frequently the causative agents of outbreaks.
The authors acknowledge the financial support from the Natural Sciences and Engineering Research Council of Canada (NSERC) and Fonds Québécois de la Recherche sur la Nature et les Technologies (FQRNT). We also thank Mrs March from the Newfoundland and Labrador National Public Health Laboratory for providing the norovirus samples and Mr Pierre-Antoine Thériault for his technical assistance.