SEARCH

SEARCH BY CITATION

Keywords:

  • murine norovirus;
  • norovirus;
  • process control;
  • RT-qPCR;
  • water

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. References

Aims

To provide a rapid and sensitive method for detecting NoV GI and NoV GII in water and to evaluate the use of the murine norovirus (MNV-1) as a process control.

Methods and Results

The method is based on viral concentration by filtration on electropositive filters and direct lysis of adsorbed viruses from filters before RNA extraction and RT-qPCR amplification. An one-step multiplex RT-qPCR assay was developed for the simultaneous detection of NoV GI, NoV GII and MNV-1. Then, water samples were artificially contaminated to determine mean virus recoveries and method sensitivity. The method showed a higher sensitivity for detecting NoV GII (103 genome copies per 0·5 l) than for NoV GI (104 genome copies per 0·5 l) in the presence of MNV-1 regardless of the type of water. The data also showed that MNV-1 is a robust option as process control.

Conclusions

The method described provides a valuable tool for the monitoring of potential public health risks associated with NoV contamination in drinkable water.

Significance and Impact of Study

Given the increasing evidence for NoV involvement in food outbreaks, the one-step multiplex RT-qPCR assay we used in this study would be a very useful tool to investigate NoV contamination in other food products.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. References

Gastroenteritis is a major public health problem worldwide, and human enteric viruses are the most common agents of foodborne gastroenteritis. Noroviruses (NoV), belonging to the Caliciviridae family, are pathogens that cause acute gastroenteritis. NoV are small nonenveloped viruses and have a positive-sense, single-stranded RNA genome, organized into three open reading frames (ORF). NoV are divided into five genogroups based on phylogenetic analysis of the capsid protein (Zheng et al. 2006). Genogroup I (GI), GII, and rarely GIV, are responsible for human outbreaks. NoV infection may induce vomiting, diarrhoea, mild fever, abdominal cramping and nausea in infected individuals. Moreover, noroviruses have recently been associated with multiple clinical outcomes other than gastroenteritis (Karst 2010). NoV are characterized by high environmental stability, and only few infectious viral particles are necessary to induce disease (Teunis et al. 2008). Transmission of these highly infectious viruses occurs mainly via the faecal–oral route, by ingestion of contaminated water and food, particularly shellfish, soft fruits and vegetables, but also through person-to-person contact and exposure to fomites (Beuchat 2006; Hewitt et al. 2007; Le Guyader et al. 2009; Lopman et al. 2012; Matthews et al. 2012; Zhou et al. 2012). NoV outbreaks occur most commonly in semi-closed communities such as restaurants, nursing homes, hospitals, schools, day care centres and cruise chips (Fankhauser et al. 2002).

Over the last few years, NoV diagnostics have improved, but there is currently no reliable method available for cultivation of NoV, and detection of this pathogen in foods relies primarily on molecular methods. Although enteric viruses have been detected in a range of samples, the main obstacles to routine detection of NoV in food and water samples are the low levels of virus contamination, variability in virus or nucleic acid extraction, the presence of substances that inhibit molecular detection and NoV genetic variability (Stals et al. 2012). Viral concentration methods in water include adsorption-elution methods using negatively or positively charged membrane filters, viral extraction and detection by real-time RT-PCR assays (Huguet et al. 2012).

Even if ISO/TS 15216-1 and 15216-2 methods have been recently published for a range of risk foods including soft fruits, bottled water and vegetables, they need to be further validated before publication as ISO/CEN standard methods. As recommended by ISO/TS 15216-1 and 15216-2, primers and probes for the individual quantification/detection of NoV GI and NoV GII should target the conserved ORF1/ORF2 junction of the genome. Another general requirement for viral diagnosis is the use of a process control to monitor the efficiency of the concentration of viral particles and the extraction of nucleic acid (Lees 2010). Although the TAG proposed the MC0 strain of Mengo virus (Costafreda et al. 2006; Le Guyader et al. 2009), there is as yet no consensus on the choice of process control (Lees 2010). The selected virus should exhibit similar morphological and physicochemical properties and environmental persistence to the target viruses, thus providing comparable extraction efficiency (Lees 2010). Ideally, the process control should be unlikely to naturally contaminate the tested food sample (Stals et al. 2012).

The first murine norovirus (MNV) was recently characterized and adapted to cell culture on murine macrophage-related cells (Karst et al. 2003; Wobus et al. 2004). MNV is morphologically and genetically similar to human noroviruses and shows considerable promise as a human norovirus surrogate (Wobus et al. 2006). On one hand, MNV-1 has been successfully tested as a process control when detecting NoV and HAV in some food samples (Stals et al. 2011a,b; Martin-Latil et al. 2012a) and HEV in bottled water (Martin-Latil et al. 2012b). On the other hand, the published method based on filtration of water and direct lysis of viruses on membrane has been found to be robust and may be useful and efficient for routine analyses of enteric viruses in bottled water (Martin-Latil et al. 2012ab; Perelle et al. 2009; Schultz et al. 2011).

With the aim of extending the use of a single process control for the detection of the main enteric viruses in food and water samples, the use of MNV-1 as a process control was evaluated for detecting NoV GI and NoV GII in bottled and tap water using filtration followed by a direct elution-based method. For this purpose, an one-step multiplex RT-qPCR was developed to detect NoV GI, NoV GII and MNV-1 simultaneously.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. References

Viruses and cells

NoV GI and NoV GII were obtained from stool samples (E3041 and E2694, respectively) of infected humans provided by the French National Reference Centre for Enteric Viruses in Dijon, France. The faecal samples were suspended in 10 mmol l−1 phosphate-buffered saline (PBS), pH 7·4, to obtain a final 10% suspension (w/v), vortexed and centrifuged at 4000 g for 20 min at 4°C. Aliquots of 100 μl were kept frozen at −80°C for later use. The titres of the clarified faecal suspensions were established in NoV GI and NoV GII genomic copies with an RT-qPCR standard curve obtained with the tenfold-diluted in vitro RNA transcripts.

Dr. H. Virgin from Washington University in the USA provided ANSES's Fougères Laboratory in France with MNV-1 (CW1 strain) which was then propagated in mouse leukaemic monocyte macrophage (RAW 264.7, ATCC TIB-71) cell line (Cannon et al. 2006). RAW 264.7 was grown at 37°C in an atmosphere containing 5% CO2 in Dulbecco's modified Eagle's medium (DMEM) supplemented with 1% L-glutamine, 1% nonessential amino acids, 10% foetal bovine serum (HyClone; Invitrogen, Cergy Pontoise, France) and 0·5% penicillin–streptomycin. MNV-1 stock containing 5·62 × 106 TCID50 per ml was produced by ANSES's Fougères Laboratory in France as previously described (Wobus et al. 2004).

NoV NoV RNA transcripts

The NoV GI cDNA and NoV GII cDNA corresponding, respectively, to the 5257–5413 and 4981–5135 positions of the genomic sequence (M87661) and (X86557) were cloned in pGEM-T Easy vector (Promega, Charbonnières-les-Bains, France) and propagated in E. coli One Shot® TOP10F' (Invitrogen). High-quality DNA plasmids containing NoV regions were purified using the Qiagen Plasmid midi kit (Qiagen, Courtaboeuf, France) according to the manufacturer's protocol. Then, the NoV GI and NoV GII DNA plasmids were digested with NcoI (Invitrogen) and SpeI (Invitrogen), respectively, and transcripts were obtained using the MEGAscript® kit (Ambion, Fisher Scientific, Illkirch, France) according to the manufacturer's protocol. Synthesized RNA was treated twice with RNase-free DNase according to the manufacturer's protocol to remove the DNA template following transcription and purified using the RNeasy Mini kit (Qiagen). The synthesized RNA was confirmed with RT-qPCR and quantified by measuring absorbance with the NanoDrop 1000 spectrophotometer (Thermoscientific, Courtaboeuf) at 260 nm. Based on this approach, the clarified suspension stocks of NoV GI and NoV GII had titres of approximately 1·21 × 109 genome copies/ml and 1·34 × 107 genome copies/ml, respectively.

NoV NoV GI, NoV GII and MNV-1 inocula for developing multiplex real-time RT-PCR assays

For individual detection, serial 10-fold dilutions of viral inoculum and RNA extraction were performed and lead to virus quantities per RT-qPCR assay ranging from 9 × 10−1 to 9 × 105 genome copies for NoV GI, from 10−1 to 104 genome copies for NoV GII and from 4 × 10−3 to 4 × 103 TCID50 for MNV-1.

For simultaneous detection, each of the five inoculation levels of NoV GI was mixed with each of the five inoculation levels of NoV GII, producing 24 samples including only one negative control. A fixed amount of 550 TCID50 of MNV-1 per sample was added to these 24 samples before RNA extraction. In conclusion, the 24 RNA extracts allowed the testing of 40 TCID50 of MNV-1 plus NoV GI among rounded 0, 9, 9 × 102, 9 × 104 and 9 × 105 genome copies and NoV GII among rounded 104, 103, 102, 10 and 0 genome copies per RT-qPCR assay (Table 1).

Table 1. The effects of NoV GII genome quantities on NoV GI detection in a multiplex real-time PCR assay and vice versa. Different genome copy numbers (0, 9, 9 × 102, 9 × 104 and 9 × 105 copies) of NoV GI are combined with different genome copy numbers (104, 103, 102, 10 and 0 copies) of NoV GII with a fixed amount of process control MNV-1 (550 TCID50). Table 1A and 1B show Ct mean values obtained for NoV GI and NoV GII, respectively, and Ct shifts calculated for each level of inoculation. Results are expressed as Ct mean values ± standard deviation (SD) from three repeated experiments
(A)
NoV GI detection (copies/RT-qPCR assay)NoV GII (copies/RT-qPCR assay)Ct-shift ranging
010102103104
939·62ND40·72 ± 0·30NDND 
9 × 10234·59 ± 0·2734·29 ± 0·7834·59 ± 0·7534·74 ± 1·0735·27 ± 0·020·01–0·68
9 × 10428·45 ± 1·3027·90 ± 0·8627·33 ± 1·1527·72 ± 1·1127·64 ± 2,000·56–1·12
9 × 10524·70 ± 1·0224·97 ± 1·0724·91 ± 1·2625·09 ± 1·2825·91 ± 1·460·21–1·21
(B)
NoV GII detection (copies/RT-qPCR assay)NoV GI (copies/RT-qPCR assay)Ct-shift ranging
099 × 1029 × 1049 × 105
1037·00 ±  0·2436·64 ± 1·2836·84 ± 0·7637·53 ± 1·0437·72 ± 0·500·1–0·72
10233·12 ± 0·3133·16 ± 0·8333·35 ± 0·7832·90 ± 0·5133·82 ± 0·420·04–0·70
10329·42 ± 0·6729·50 ± 0·5829·83 ± 0·5829·39 ± 0·7829·74 ± 0·430·03–0·42
10426·66 ± 0·6126·99 ± 0·6427·09 ± 1·0326·59 ± 0·4626·96 ± 0·560·08–0·43

Inoculation of water samples

The principal chemical characteristics of the purchased bottled mineral water used to perform the study were as follows: Ca2+, 11·5 mg l−1; Mg2+, 8 mg l−1; Na2+,11·6 mg l−1; K+, 6·2 mg l−1; Cl,13·5 mg l−1; inline image, 6·3 mg l−1; inline image, 8·1 mg l−1; SiO2, 31·7 mg l−1, inline image, 71·0 mg l−1; F, 0·2 mg l−1. The pH was measured in the laboratory at 7. Tap water (from Maisons-Alfort) used in this study was collected into glass flasks. Chlorine residues in 0·5 l of tap water were neutralized with 250 μl of 10% sodium thiosulfate (Na2S2O3,) solution (Sigma-Aldrich, Saint-Quentin-Fallavier, France) (Mendez et al. 2004), and the tap water was stored overnight at 4°C before viral inoculation.

Drinkable volume of bottled mineral water and tap water samples (0·5 l) were spiked with 100 μl of 10-fold dilutions in DEPC-treated water (Fisher Bioblock Scientific, Illkirch, France) of clarified NoV GI and NoV GII suspension stocks. The co-inoculation of bottles of water was performed with rounded inoculation levels of 0, 102, 103, 104, 106 and 107 NoV GI genome copies, respectively, added to 0, 105, 104, 103, 102, 10 NoV GII genome copies leading to six water samples inoculated with norovirus mixtures named M1 to M6. Each experiment set comprised 6 bottles of water spiked with M1 to M6 alone and 6 bottles co-inoculated with 550 TCID50 MNV-1 as a process control. The spiked bottle samples were homogenized using ten hand inversions to shake. Uninoculated water samples were used as a negative control. Each experiment set, from spiking to RNA extraction, was performed three times for each type of water, and RNA extracts (neat and diluted ten-fold) were analysed in duplicate with the multiplex RT-qPCR assay.

Sample processing for recovery of viruses and viral RNA extraction

The protocol used for recovery of viruses and viral RNA extraction was as previously described from drinkable volume of water (0·5–1·5 l) (Perelle et al. 2009; Schultz et al. 2011; Martin-Latil et al.2012a, 2012b). Viruses were concentrated from 0·5 l of inoculated water samples by membrane filtration under vacuum using a Zetapor (Cuno Filtration SAS 3M, Cergy Pontoise, France) 47-mm positively charged membrane of pore size 0·45 μm. The filter membrane was then directly incubated for 10 min at room temperature in a 60-mm-diameter Petri dish containing 3 ml of NucliSens® easyMAG™ lysis buffer (Biomérieux, Marcy l'Etoile, France) and was then subjected to the NucliSens® easyMAG™ platform (Biomérieux) for total nucleic acid extraction by the ‘off-board Specific A’ protocol according to manufacturer's instructions. Nucleic acids were finally eluted in 70 μl of elution buffer and stored at −80°C.

Primers and probes

Primers and probes for the individual quantification of NoV GI and NoV GII were previously described (Kageyama et al. 2003; Loisy et al. 2005; da Silva et al. 2007; Svraka et al. 2007) and were chosen by the CEN/TC/WG6/TAG4 research group. The primers and the TaqMan® probe targeting the ORF1 of the murine norovirus were designed using Beacon Designer software (Bio-Rad, Marnes-la-Coquette, France). The NoV GI, NoV GII and MNV-1 probes were, respectively, labelled with the 6-FAM, ROX and CY5 reporter dyes at the 5′-end and a BHQ1, BHQ2 or BBQ650 at the 3′-end.

The sequence of the primer pairs and the TaqMan probes used was as follows: For NoV GI, the sense primer (QNIF4) was 5′-CGCTGGATGCGNTTCCAT-3′, the antisense primer (NV1LCR) was 5′-CCTTAGACGCCATCATCATTTAC-3′ and the TaqMan probe (NVGG1p) was 5′-FAM-TGGACAGGAGAYCGCRATCT-BHQ1-3′. For NoV GII, the sense primer (QNIF2) was 5′-ATGTTCAGRTGGATGAGRTTCTCWGA-3′, the antisense primer (COG2R) was 5′-TCGACGCCATCTTCATTCACA -3′ and the TaqMan probe (QNIFS) was 5′-ROX-AGCACGTGGGAGGGCGATCG-BHQ2-3′. For MNV-1, the sense primer (MNV-3193-F) was 5′-CCGCCATGGTCCTGGAGAATG-3′, the antisense primer (MNV-3308R) was 5′-GCACAACGGCACTACCAATCTTG-3′ and the TaqMan probe (MNV-3227-T) was 5′-CY5–CGTCGTCGCCTCGGTCCTTGTCAA-BBQ650-3′. All the primers and probes were purchased from Eurofins (Les Ulis, France).

RT-qPCR conditions

One-step multiplex RT-qPCR amplifications were performed in duplicate on the CFX96™ real-time PCR detection system from Bio-Rad.

Reactions were performed in a 25 μl reaction mixture containing 1 × of RNA UltraSense™ master mix and 1·25 μl of RNA UltraSense™ enzyme mix, which are components of the RNA UltraSense™ One-Step Quantitative RT-PCR System (Fisher Scientific, Illkirch, France), 2 U RNase inhibitor (Applied Biosystems), 5 μg of bovine serum albumin (Ambion), 500 nmol l−1 of NoV GI and NoV GII forward primer, 900 nmol l−1 of NoV GI and NoV GII reverse primer, 250 nmol l−1 of NoV GI and NoV GII probe, 600 nmol l−1 of MNV-3193-F forward primer, 600 nmol l−1 of MNV-3308R reverse primer, 250 nmol l−1 of MNV-3227-T probe and 5 μl of RNA extract. Positive controls containing RNA extracted from virus suspensions and a negative control containing all the reagents except the RNA template were included with each set of reaction mixtures. The one-step RT-qPCR program involved 60 min reverse transcription of RNA at 55°C, followed by a 15 min denaturation step at 95°C and lastly 40 cycles of 15 s at 95°C, 1 min at 60°C and 1 min at 65°C. Fluorescence was recorded by the apparatus at the end of the elongation steps (1 min at 65°C) for each amplification cycle. All samples were characterized by a corresponding Ct value. Negative samples gave no Ct value. Standard curves for NoV GI, NoV GII, MNV-1 were generated with 10-fold diluted RNA extracts resulting from viral stock suspension in distilled water or 10-fold diluted RNA transcripts for noroviruses. The slopes (S) of the regression lines were used to calculate the amplification efficiency (E) of the RT-qPCRs, according to the formula E = 10|−1/s| –1 to determine the performance of RT-qPCR assays. NoV GI, NoV GII and MNV-1 recovery rate percentages in spiked samples were calculated with reference to the corresponding standard curve and the following formula:

  • display math

Statistical analysis

The impact of the presence of MNV-1 on extraction yields of NoV GI and NoV GII has been statistically evaluated using an one-way analysis of variance (Anova) with the Anovan subroutine from the statistics toolbox of MATLAB software (version 6.5.1; The Math Works Inc., Natick, MA, USA). Four factors for extraction yield of NoV GI or NoV GII were further studied: (i) norovirus mixtures (M4, M5 and M6 for NoV GI and M2, M3, M4 and M5 for NoV GII), (ii) dilution of the sample (nondiluted/diluted tenfold), (iii) interassay variability and (iv) water (bottled water/tap water). The influence of three factors on the extraction yield of MNV-1 (dilution, interassay variability, water) was investigated in the same way. For this, we used multiway analysis of variance with the anovan subroutine from the statistics toolbox of MATLAB software (version 6.5.1; The Math Works Inc., Natick, MA, USA).

To test for differences between levels of each factor, the multiple comparison procedure was applied (MATLAB's multcompare subroutine). Multiple comparison methods are designed to provide an upper bound on the probability that any comparison will be incorrectly found significant. Population marginal means were also computed, with the effect of specific factors on the means removed. In the multcompare subroutine, we used critical values from Scheffé's S procedure, derived from the F distribution. This procedure provides a simultaneous confidence level for comparisons of all linear combinations of the means.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. References

Development of a multiplex RT-qPCR assay for the simultaneous detection of NoV GI, NoV GII and MNV-1

Singleplex vs multiplex RT-qPCR assays for individual NoV GI, NoV GII and MNV-1 detection

RNA extracts from 10-fold dilutions of viral stock suspension for each viral target were used to compare the standard curves obtained with the RT-qPCR assays using a singleplex and multiplex mix in three experiments realized in single. The parameters of the standard curves of the individual NoV GI, NoV GII and MNV-1 reactions showed a mean RT-qPCR efficiency of 109, 118 and 99%, respectively, within the singleplex RT-qPCR mix. Individual reactions within the multiplex RT-qPCR mix were as efficient, as they showed a mean RT-qPCR efficiency of 104, 106 and 98% for NoV GI, NoV GII and MNV-1, respectively. The limit of detection (LOD) has been defined as the lowest amount of NoV detected in the three experiments, consequently, LOD occurred with 3/3 positive Ct determinations. The LOD was 90 and 10 genome copies per assay for NoV GI and NoV GII, respectively, regardless of the singleplex or multiplex RT-qPCR assays used (data not shown). These results underlined the highest sensitivity for detecting NoV GII and showed the reliability of the multiplex RT-qPCR assay for the detection of NoV GI, NoV GII and MNV-1.

Multiplex RT-qPCR for the simultaneous detection of NoV GI, NoV GII and MNV-1

The ability of the multiplex assay to detect both target viruses NoV GI and NoV GII, and MNV-1 used as a process control was further assessed. The 24 RNA samples combining various amounts of NoV GI and NoV GII and a fixed amount of MNV-1 were tested with the multiplex RT-qPCR assay. Reproducible detection of the process control in the range of 22–24 Ct by RT-qPCR (mean Ct value of 23·06 ± 0·64) was obtained regardless of the quantities of NoV GI and/or NoV GII, confirming that 550 TCID50 of MNV-1 per sample is a reliable input. The resulting Ct values obtained for NoV GI and NoV GII with MNV-1 present are reported in Table 1A and 1B, respectively.

Irrespective of the NoV GII quantities tested and with MNV-1 present, each amount of NoV GI (9 × 102, 9 × 104 and 9 × 105 genome copies) was detected at the expected Ct values with a Ct shift which did not exceed 1·21. Similarly, NoV GII detection was not affected by the co-amplification of NoV GI regardless of the viral load. The four inoculation levels of NoV GII (10, 102, 103, 104) were always detected with a Ct shift which was not higher than 0·72. Ct shift does not exceed 5% of the expected Ct value showing no interference of NoVGI on NoVGII detection and vice versa. Moreover, the 95% confidence intervals of the slopes and intercepts of all regression lines were obtained for NoV GI whatever the quantity of NoV GII overlap considerably and vice versa (data not shown). These results showed that the multiplex assay was reliable for detecting NoV GI and NoV GII.

Detection of NoV GI and NoV GII in tap and bottled water samples using MNV-1 as a process control by multiplex RT-qPCR assay

Standard curves

The multiplex RT-qPCR assay was further used to detect NoV GI and NoV GII in samples of tap water and bottled mineral water spiked with M1 to M6 norovirus mixtures with or without the presence of MNV-1 as a process control. The parameters of RT-qPCR curves obtained with RNA extracts resulting from spiked water samples were shown in Table 2. As expected, no viral RNA was detected in the uninoculated water samples. The 95% confidence intervals of the slopes and intercepts of the NoV GI and NoV GII calibration dilution curves are also reported in Table 2. The confidence intervals of the slopes and intercepts of all regression lines obtained for NoV GI and NoV GII overlap considerably. These results show that neither the presence of MNV-1 nor the water sample type has affected RT-qPCR amplification of NoV regardless of the genogroup.

Table 2. Parameters of RT-qPCR amplification curves obtained for NoV detection in spiked water samples with or without the addition of MNV-1 according to the type of water
Type of waterPresence of MNV-1Slope (α)Intercept (β)Residual standard deviation (σ)
EstimateCI [2·5%; 97·5%]EstimateCI [2·5%; 97·5%]
NoV GI
Bottled waterYes−3·74[−4·21; −3·28]45·33[43·67; 46·99]1·37
No−3·59[−4·23; −2·96]44·83[42·49; 47·17]1·75
Tap waterYes−3·28[−3·85; −2·72]46·24[43·56; 48·91]1·40
No−3·03[−3·43; −2·62]44·69[42·84; 46·54]1·23
NoV GII
Bottled waterYes−3·12[−3·49; −2·75]41·39[40·30; 42·49]0·77
No−3·03[−3·28; −2·77]41·08[40·39; 41·78]0·67
Tap waterYes−3·41[−3·83; −2·98]42·84[41·64; 44·04]1·02
No−3·17[−3·59; −2·74]41·99[40·84 43·13]1·11
Limits of detection of NoV GI and NoV GII

The LOD, which was defined as the lowest amount of NoV in a test sample detected in the three sets of experiments, was determined. An assay was considered positive if at least one positive Ct determination resulted from two RT-qPCR duplicates. In both types of water, the LOD of NoV GI remained at 104 genome copies per 0·5 l in the presence of MNV-1, whereas the LOD of NoV GII increases from 102 to 103 genome copies per 0·5 l when MNV-1 process control was added. At inoculation levels of 102 genome copies per 0·5 l of water, we were able to detect NoV GII in 2/3 experiments in both types of water, showing a very slight decrease in sensitivity when the process control was added. These results also showed that NoV GII detection sensitivity is always higher than when detecting NoV GI in both types of water.

Mean virus recoveries

Virus recoveries established from spiked experiments are reported in Table 3. All experiments with water samples spiked with NoV GI, NoV GII and MNV-1 showed that the process control was detected consistently in RNA extracts. The average of NoV GI recoveries for every level of NoV GI inoculation calculated with pure RNA extracts ranged from 1·93 to 44·01% in bottled water and from 0·87 to 16·51% in tap water. In the same way, the type of water did not seem to influence NoV GII recovery as the percentages ranged from 2·48 to 16·37% in bottled water and from 1·57 to 16·40% in tap water. We observed that recovery rates of NoV GI and NoV GII increased when the levels of inocula were below the LOD, suggesting that they influenced the respective recovery success rates. The recovery efficiencies of NoV GI and NoV GII from spiked bottled water and tap water samples were not significantly influenced by the presence of MNV-1 (one-way anova; P = 0·057 for GI and P = 0·5088 for GII). Therefore, the averages of the recovery rates of NoV GI and NoV GII for inoculum levels higher or equal to the LOD were calculated without taking into account the presence of MNV-1. In bottled water, the mean recovery rates of NoV GI and NoV GII calculated with undiluted RNA extracts (3·91 and 4·30%, respectively) were almost the same as those calculated with diluted RNA extracts (3·77 and 4·30%, respectively). However, in tap water, the mean recovery rate of NoV GI increased with the dilution of RNA extracts from 1·45 to 5·88%, whereas the mean recovery rate of NoV GII did not change (4·06% vs 4·25%). These results suggest that the presence of PCR inhibitors in tap water only influence NoV GI.

Table 3. Mean percentage recovery calculated for NoV and MNV-1 in bottled water and tap water. For each norovirus mixtures (M1 to M6), three experiments were performed and RNA extracts were tested twice, resulting in six cycle threshold (Ct) values for each sample type. An experiment set with fifty percentage of positive Ct values was considered to be a positive assay. The number of positive Ct values for NoV is mentioned in brackets. The results corresponding to the LOD values for each sample type are shown in bold
SamplesInocula (per 0·5l) Bottled water recovery (%)Tap water recovery (%)
NoV GI (genome copies)NoV GII (genome copies)MNV-1 (TCID50)NoV GI (genome copies)NoV GII (genome copies)MNV-1 (TCID50)NoV GI (genome copies)NoV GII (genome copies)MNV-1 (TCID50)
  1. nd, not detected; Neat: nondiluted RNA extract; diluted: 10-fold diluted RNA extract.

M10Neatndndndndndnd
M21021050Neat13·83 ± 12·13 (2/6)3·1 ± 0·73 (6/6)ndnd2·80 ± 1·39 (6/6)nd
Dilutednd5·88 ± 1·81 (6/6) nd3·85 ± 1·94 (6/6) 
M31031040Neat44·01 ± 50·90(2/6)3·81 ± 1·39 (6/6)nd16·51 ± 20·19(2/6)1·80 ± 0·43 (6/6)nd
Dilutednd4·26 ± 2·52(6/6) 35·90 (1/6)2·73 ± 0·79 (6/6) 
M41041030Neat 4·81 ± 2·81 (6/6) 5·86 ± 2·36 (6/6)nd 1·92 ± 0·83 (6/6) 1·57 ± 1·40 (6/6)nd
Diluted6·32 ± 7·70 (3/6)2·4 ± 1·72 (6/6) 7·6 ± 5·98 (3/6)3·22 ± 1·88 (6/6) 
M51061020Neat3·89 ± 1·76 (6/6) 16·37 ± 12·15(6/6) nd1·25 ± 0·51 (6/6) 5·15 ± 3·29(6/6) nd
Diluted3·91 ± 2·75 (5/6)42·1 (1/6) 4·86 ± 0·74 (5/6)nd 
M6107100Neat3·14 ± 2·61 (6/6)ndnd1·48 ± 0·57 (6/6)ndnd
Diluted2·69 ± 2·13 (6/6)nd 4·57 ± 2·13 (6/6)nd 
M1500Neatndnd7·31 ± 3·33ndnd1·56 ± 0·51
M2102105500Neat9·38 ± 3·19 (2/6)2·48 ± 0·30 (6/6)2·1 ± 0·69nd3·12 ± 0·84 (6/6)1·39 ± 0·27
Dilutednd4·23 ± 0·88 (6/6)6·73 ± 3·54nd5·28 ± 3·08 (6/6)3·37 ± 1·39
M3103104500Neat12·09 ± 9·40 (3/6)5·41 ± 1·15 (6/6)7·03 ± 2·94nd5·41 ± 7·11 (6/6)1·32 ± 0·54
Dilutednd5·53 ± 2·05(6/6)10·32 ± 4·5848·06 ± 53·20 (2/6)2·01 ± 1·34 (6/6)2·01 ± 0·42
M4104103500Neat 6·41 ± 2·81 (6/6) 5·12 ± 1·56 (6/6) 5·61 ± 2·40 1·23 ± 0·6 (6/6) 9·66 ± 13·91(6/6) 1·79 ± 0·94
Diluted3·56 ± 2·10 (6/6)3·52 ± 1·00 (5/6)9·3 ± 3·524·96 ± 4·25 (3/6)8·45 ± 6·98 (4/6)3·6 ± 1·65
M5106102500Neat3·31 ± 0·70 (6/6)2·58 ± 0·01 (2/6)4·73 ± 1·041·95 ± 0·55 (6/6)16·40 ± 16·00 (4/6)2·25 ± 0·63
Diluted3·85 ± 1·21(6/6)nd5·9 ± 3·313·10 ± 1·52 (6/6)16·84 (1/6)6·91 ± 6·09
M610710500Neat1·93 ± 0·87 (6/6)nd3·41 ± 1·290·87 ± 0·28 (6/6)nd1·53 ± 0·78
Diluted2·23 ± 0·74 (6/6)nd4·07 ± 0·82·21 ± 0·47 (6/6)nd2·58 ± 0·82

We also observed that the mean recovery rate of the MNV-1 process control increased with the dilution of RNA extracts from 5·03 to 7·26% for bottled water and from 1·64 to 3·69% for tap water.

Influence of experimental factors on extraction yield of NoV GI and NoV GII and MNV-1

The extraction recoveries obtained for NoV GI, NoV GII and MNV-1 from spiking experiments were statistically analysed to assess the influence of the different experimental factors. The results of the multiway anovas and the multiple comparison tests are displayed in Figs 1, 2 and 3 for NoV GI, NoV GII and MNV-1, respectively. The standard anovas present variability due to the differences between the levels of each factor accounted for in the model, while the remaining variability is not explained by any systematic source. These figures give information on the difference between levels taken by the different factors. Two means are significantly different (at the 0·05 level) if their intervals are disjoint and are not significantly different if their intervals overlap.

image

Figure 1. Population marginal means with standard error of extraction yields for NoV GI under different conditions. Two means are significantly different if their intervals are disjoint and are not significantly different if their intervals overlap. The influence of four experimental factors on NoV GI extraction was illustrated. (a) Mixtures of NoV GI and GII (P-values = 0·0002) (M4 = 104 NoV GI and 103 NoV GII; M5 = 106 NoV GI and 102 NoV GII; M6 = 107 NoV GI and 10 NoV GII), (b) dilution factor (P-values = 0·0007), (c) reproducibility (P-values = 0·060), (d) type of water (P-values = 0·057).

Download figure to PowerPoint

image

Figure 2. Population marginal means with standard error of extraction yields for NoV GII under different conditions. Two means are significantly different if their intervals are disjoint and are not significantly different if their intervals overlap. The influence of four experimental factors on NoV GII extraction was illustrated. (a) Mixtures of NoV GI and GII (P-values < 0·0001) (M2 = 102 NoV GI and 105 NoV GII; M3 = 103 NoV GI and 104 NoV GII; M4 = 104 NoV GI and 103 NoV GII; M5 = 106 NoV GI and 102 NoV GII), (b) dilution factor (P-values = 0·3193), (c) reproducibility (P-values = 0·0002), (d) type of water (P-values = 0·2405).

Download figure to PowerPoint

image

Figure 3. Population marginal means with standard error of extraction yield for MNV-1 under different conditions. Two means are significantly different if their intervals are disjoint and are not significantly different if their intervals overlap. The influence of three experimental factors on MNV-1 extraction was illustrated. (a) Dilution factor (P-values ≤ 0·0001), (b) reproducibility (P-values = 0·4056), (c) type of water (P-values ≤ 0·0001).

Download figure to PowerPoint

The multiway anovas and the multiple comparison tests showed that the differences between reproductions of experiments are only significant for the extraction yield of NoV GII. It also showed that the mixtures of NoV which were chosen for the statistical analysis (i.e. containing quantities of NoV GI and NoV GII higher or equal to their LOD) were a significant factor for NoV GI and NoV GII extraction yields. More accurately, mixture 4 containing the amount of NoV GI corresponding to its LOD was significantly different from mixtures 5 and 6. In the same way, mixture 5 containing the amount of NoV GII corresponding to its LOD for experiments performed without MNV-1 was significantly different from mixtures 2, 3 and 4. These statistical data confirmed that recovery rates of NoV GI and NoV GII increased for quantities closer to the LOD, as we previously observed. While NoV GII extraction yield was not improved by dilution of RNA samples before multiplex RT-qPCR, the ratio of marginal means showed that NoV GI and MNV-1 recoveries were slightly improved by a factor of 1·5 and 1·6, respectively. Unlike NoV GI and GII, the type of water influenced extraction yields of MNV-1 which was 2·4 times higher in bottled water than in tap water.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. References

Noroviruses are considered to be the leading cause of foodborne and waterborne disease outbreaks and acute nonbacterial gastroenteritis worldwide, with transmission from food and water or from person to person via the faecal–oral route affecting adults and children all over the world (Atmar and Estes 2006; Koo et al. 2010). Indeed, NoV cause at least 95% of all nonbacterial gastroenteritis outbreaks and 50% of all gastroenteritis outbreaks (Karst 2010). Moreover, waterborne transmission is a significant route of exposure, as contaminated water also serves as a vehicle for outbreaks of food poisoning by foods such as vegetables and shellfish. Therefore, the availability of a sensitive, rapid and efficient virus concentration method for direct detection of viruses in drinkable water is an essential tool for the prevention of waterborne viral diseases.

The method used for virus extraction from water samples was previously described and evaluated through a collaborative trial for the detection of HAV, NoV and FCV in bottled water (Schultz et al. 2011). This method has been preferred to the ISO/TS 15216 method, as the authors clearly showed its efficiency, reproducibility and robustness across laboratories. This method is a two-step procedure: viruses are concentrated using positive-charged membranes, and the adsorbed viruses are directly lysed on membrane followed by application of automatic RNA extraction equipment. Multiple pathogen detection represents an option to decrease reagent costs and speed up the detection of several pathogens in a single reaction. Therefore, in the current study, a quantitative one-step sensitive multiplex RT-qPCR assay was first characterized for the simultaneous detection of NoV (GI and GII) and MNV-1 chosen as a process control, to be further used to detect and quantify NoV and MNV-1 in water samples.

We used primers and hydrolysis probes recommended for NoV detection by the European Committee for Standardization. These primer sets were recently shown to be among the most sensitive primer sets for NoV GI and NoV GII detection (Tong et al. 2011). With the one-step singleplex and multiplex RT-qPCR developed in this study, we also showed a higher sensitivity (about 1 log10) for detecting NoV GII than for detecting NoV GI. The higher sensitivity for NoV GII detection could be explained by the molecular model by itself. However, Stals et al. (2009) have developed a quantitative two-step multiplex real-time reverse transcriptase PCR assay to detect NoV and MNV-1, which was as sensitive for NoV GI as for NoV GII (LOD of 10 genome copies per assay). This higher sensitivity for NoV GI detection may be explained by the fact that a two-step RT-qPCR method could be more efficient and sensitive than an one-step method. However, we preferred an one-step protocol to minimize handling and therefore reduce chances of pipetting errors and cross-contamination. Moreover, unlike the results obtained by Stals et al. (2009), there was no significant competitive effect between the NoV GI and NoV GII reactions. In this study, the multiplex assay was found reliable for detecting NoV GI and NoV GII within the same sample based on the range of Ct shifts and parameters of regression lines.

The higher sensitivity for detecting NoV GII than for NoV GI has been also found to detect NoV regardless of the type of water. We also noticed an influence of the inoculum level on the recovery efficiency of NoV GI and NoV GII as previously shown by Stals et al. (2011a,b) on frozen raspberries and by Fumian et al. (2009) on lettuce. In addition, the variability between experiments was observed for inoculation levels equal and/or below the LOD where recovery rates were the highest. This interassay variability was significant for NoV GII detection and has also been observed for the detection of various micro-organisms in food (Le Guyader et al. 2009; Postollec et al. 2011; Martin-Latil et al. 2012a). We found significant improvement in NoV GI extraction yields calculated with diluted RNA samples (factor 1·5), which was not observed in the case of NoV GII detection. Direct lysis of viral particles on filtration membranes should limit loss of viral particles, but inhibitors may be also concentrated (Gregory et al. 2006; da Silva et al. 2007), and their effects on PCR amplification may vary according to viral targets. Nevertheless, the LOD of the method for detecting NoV was not improved after sample dilution. Therefore, at very low levels of contamination, dilution itself probably brings the sample below levels of detection, with a very limited impact of inhibitory substances for the detection of NoV in water.

Unlike the norovirus target, the type of water had a significant effect on the recovery rates of MNV-1 used as a control process, as the ratio of marginal means was 2·4 times higher in bottled water than in tap water. The influence of water type has already been observed in the detection of enteric viruses such as HAV, HEV and poliovirus (Blaise-Boisseau et al. 2010; Huguet et al. 2012; Martin-Latil et al. 2012b). However, recovery rates obtained for NoV extraction were quite similar regardless of the water type tested, in agreement with Huguet et al. (2012), showing that tap water quality did not affect the analytical performance of molecular methods for detecting NoV.

As shown in our previous study, the inhibition of MNV-1 amplification was slightly more pronounced in tap water than in bottled water (Martin-Latil et al. 2012b). Overall, it would appear that NoV are detected consistently with the same range of recovery rates of MNV-1, showing that MNV-1 is a robust option for routine sample processes in tap and bottled water analysis and should be further tested for other type of water.

In conclusion, the method described provides a valuable tool for the monitoring of potential public health risks associated with NoV contamination in drinkable water. It could be further interesting to determine the recovery efficiencies of NoV obtained using the described method throughout the different types of environmental water. Given the increasing evidence for NoV involvement in food outbreaks, the one-step triplex we used in this study would be a very useful tool to investigate NoV contamination in other food products.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. References

We thank Pr P. Pothier for providing NoV-contaminated stools and P. Maris for providing MNV-1. We thank Ghislaine Merle for her participation in the recombinant NoV plasmid constructions.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. References
  • Atmar, R.L. and Estes, M.K. (2006) The epidemiologic and clinical importance of norovirus infection. Gastroenterol Clin North Am 35, 275290.
  • Beuchat, L.R. (2006) Vectors and conditions for preharvest contamination of fruits and vegetables capable of causing enteric diseases. Br Food J 108, 3853.
  • Blaise-Boisseau, S., Hennechart-Collette, C., Guillier, L. and Perelle, S. (2010) Duplex real-time qRT-PCR for the detection of hepatitis A virus in water and raspberries using the MS2 bacteriophage as a process control. J Virol Methods 166, 4853.
  • Cannon, J.L., Papafragkou, E., Park, G.W., Osborne, J., Jaykus, L.A. and Vinje, J. (2006) Surrogates for the study of norovirus stability and inactivation in the environment: a comparison of murine norovirus and feline calicivirus. J Food Prot 69, 27612765.
  • Costafreda, M.I., Bosch, A. and Pinto, R.M. (2006) Development, evaluation, and standardization of a real-time TaqMan reverse transcription-PCR assay for quantification of hepatitis A virus in clinical and shellfish samples. Appl Environ Microbiol 72, 38463855.
  • Fankhauser, R.L., Monroe, S.S., Noel, J.S., Humphrey, C.D., Bresee, J.S., Parashar, U.D., Ando, T. and Glass, R.I. (2002) Epidemiologic and molecular trends of “Norwalk-like viruses” associated with outbreaks of gastroenteritis in the United States. J Infect Dis 186, 17.
  • Fumian, T.M., Leite, J.P., Marin, V.A. and Miagostovich, M.P. (2009) A rapid procedure for detecting noroviruses from cheese and fresh lettuce. J Virol Methods 155, 3943.
  • Gregory, J.B., Litaker, R.W. and Noble, R.T. (2006) Rapid one-step quantitative reverse transcriptase PCR assay with competitive internal positive control for detection of enteroviruses in environmental samples. Appl Environ Microbiol 72, 39603967.
  • Hewitt, J., Bell, D., Simmons, G.C., Rivera-Aban, M., Wolf, S. and Greening, G.E. (2007) Gastroenteritis outbreak caused by waterborne norovirus at a New Zealand ski resort. Appl Environ Microbiol 73, 78537857.
  • Huguet, L., Carteret, C. and Gantzer, C. (2012) A comparison of different concentration methods for the detection of viruses present in bottled waters and those adsorbed to water bottle surfaces. J Virol Methods 181, 1824.
  • Kageyama, T., Kojima, S., Shinohara, M., Uchida, K., Fukushi, S., Hoshino, F.B., Takeda, N. and Katayama, K. (2003) Broadly reactive and highly sensitive assay for Norwalk-like viruses based on real-time quantitative reverse transcription-PCR. J Clin Microbiol 41, 15481557.
  • Karst, S.M. (2010) Pathogenesis of noroviruses, emerging RNA viruses. Viruses 2, 748781.
  • Karst, S.M., Wobus, C.E., Lay, M., Davidson, J. and Virgin, H.W.t. (2003) STAT1-dependent innate immunity to a Norwalk-like virus. Science 299, 15751578.
  • Koo, H.L., Ajami, N., Atmar, R.L. and DuPont, H.L. (2010) Noroviruses: the leading cause of gastroenteritis worldwide. Discov Med 10, 6170.
  • Le Guyader, F.S., Parnaudeau, S., Schaeffer, J., Bosch, A., Loisy, F., Pommepuy, M. and Atmar, R.L. (2009) Detection and quantification of noroviruses in shellfish. Appl Environ Microbiol 75, 618624.
  • Lees, D. and CEN WG6 TAG4. (2010) International standardization of a method for detection of human pathogenic viruses in molluscan shellfish. Food Environ Virol 2, 146155.
  • Loisy, F., Atmar, R.L., Guillon, P., Le Cann, P., Pommepuy, M. and Le Guyader, F.S. (2005) Real-time RT-PCR for norovirus screening in shellfish. J Virol Methods 123, 17.
  • Lopman, B., Gastanaduy, P., Park, G.W., Hall, A.J., Parashar, U.D. and Vinje, J. (2012) Environmental transmission of norovirus gastroenteritis. Curr Opin Virol 2, 96102.
  • Martin-Latil, S., Hennechart-Collette, C., Guillier, L. and Perelle, S. (2012a) Comparison of two extraction methods for the detection of hepatitis A virus in semi-dried tomatoes and murine norovirus as a process control by duplex RT-qPCR. Food Microbiol 31, 246253.
  • Martin-Latil, S., Hennechart-Collette, C., Guillier, L. and Perelle, S. (2012b) Duplex RT-qPCR for the detection of hepatitis E virus in water, using a process control. Int J Food Microbiol 157, 167173.
  • Matthews, J.E., Dickey, B.W., Miller, R.D., Felzer, J.R., Dawson, B.P., Lee, A.S., Rocks, J.J., Kiel, J. et al. (2012) The epidemiology of published norovirus outbreaks: a review of risk factors associated with attack rate and genogroup. Epidemiol Infect 140, 11611172.
  • Mendez, J., Audicana, A., Isern, A., Llaneza, J., Moreno, B., Tarancon, M.L., Jofre, J. and Lucena, F. (2004) Standardised evaluation of the performance of a simple membrane filtration-elution method to concentrate bacteriophages from drinking water. J Virol Methods 117, 1925.
  • Perelle, S., Cavellini, L., Burger, C., Blaise-Boisseau, S., Hennechart-Collette, C., Merle, G. and Fach, P. (2009) Use of a robotic RNA purification protocol based on the NucliSens easyMAG for real-time RT-PCR detection of hepatitis A virus in bottled water. J Virol Methods 157, 8083.
  • Postollec, F., Falentin, H., Pavan, S., Combrisson, J. and Sohier, D. (2011) Recent advances in quantitative PCR (qPCR) applications in food microbiology. Food Microbiol 28, 848861.
  • Schultz, A.C., Perelle, S., Di Pasquale, S., Kovac, K., De Medici, D., Fach, P., Sommer, H.M. and Hoorfar, J. (2011) Collaborative validation of a rapid method for efficient virus concentration in bottled water. Int J Food Microbiol 145(Suppl 1), S158S166.
  • da Silva, A.K., Le Saux, J.C., Parnaudeau, S., Pommepuy, M., Elimelech, M. and Le Guyader, F.S. (2007) Evaluation of removal of noroviruses during wastewater treatment, using real-time reverse transcription-PCR: different behaviors of genogroups I and II. Appl Environ Microbiol 73, 78917897.
  • Stals, A., Baert, L., Botteldoorn, N., Werbrouck, H., Herman, L., Uyttendaele, M. and Van Coillie, E. (2009) Multiplex real-time RT-PCR for simultaneous detection of GI/GII noroviruses and murine norovirus 1. J Virol Methods 161, 247253.
  • Stals, A., Baert, L., De Keuckelaere, A., Van Coillie, E. and Uyttendaele, M. (2011a) Evaluation of a norovirus detection methodology for ready-to-eat foods. Int J Food Microbiol 145, 420425.
  • Stals, A., Baert, L., Van Coillie, E. and Uyttendaele, M. (2011b) Evaluation of a norovirus detection methodology for soft red fruits. Food Microbiol 28, 5258.
  • Stals, A., Baert, L., Van Coillie, E. and Uyttendaele, M. (2012) Extraction of food-borne viruses from food samples: a review. Int J Food Microbiol 153, 19.
  • Svraka, S., Duizer, E., Vennema, H., de Bruin, E., van der Veer, B., Dorresteijn, B. and Koopmans, M. (2007) Etiological role of viruses in outbreaks of acute gastroenteritis in The Netherlands from 1994 through 2005. J Clin Microbiol 45, 13891394.
  • Teunis, P.F., Moe, C.L., Liu, P., Miller, S.E., Lindesmith, L., Baric, R.S., Le Pendu, J. and Calderon, R.L. (2008) Norwalk virus: how infectious is it? J Med Virol 80, 14681476.
  • Tong, H.I., Connell, C., Boehm, A.B. and Lu, Y. (2011) Effective detection of human noroviruses in Hawaiian waters using enhanced RT-PCR methods. Water Res 45, 58375848.
  • Wobus, C.E., Karst, S.M., Thackray, L.B., Chang, K.O., Sosnovtsev, S.V., Belliot, G., Krug, A., Mackenzie, J.M. et al. (2004) Replication of Norovirus in cell culture reveals a tropism for dendritic cells and macrophages. PLoS Biol, 2, e432.
  • Wobus, C.E., Thackray, L.B. and Virgin, H.W.t. (2006) Murine norovirus: a model system to study norovirus biology and pathogenesis. J Virol 80, 51045112.
  • Zheng, D.P., Ando, T., Fankhauser, R.L., Beard, R.S., Glass, R.I. and Monroe, S.S. (2006) Norovirus classification and proposed strain nomenclature. Virology 346, 312323.
  • Zhou, X., Li, H., Sun, L., Mo, Y., Chen, S., Wu, X., Liang, J., Zheng, H. et al. (2012) Epidemiological and molecular analysis of a waterborne outbreak of norovirus GII.4. Epidemiol Infect 140, 22822289.